Abstract
Background
Physical activity and macronutrient intake, important contributors to energy balance, may be independently associated with female urinary incontinence (UI).
Methods
We evaluated the association of baseline self-reported physical activity and macronutrient intake, via food frequency questionnaire, with incident UI subtypes after 3 years among 19 741 postmenopausal women in the Women’s Health Initiative Observational Study. Odds ratios (ORs) for incident urgency, stress, and mixed UI were calculated using multivariable logistic regression.
Results
Women who reported total physical activity (metabolic equivalent task [MET]-hours/week) ≥30 versus <0.1 were 16% less likely to develop urgency UI (OR = 0.84; 95% CI 0.70, 1.00) and 34% less likely for mixed UI (OR = 0.66; 95% CI 0.46, 0.95), although linear trends were no longer statistically significant after adjusting for baseline weight and weight change (p trend = .15 and .16, respectively). The association between physical activity and incident stress UI was less consistent. Higher uncalibrated protein intake was associated with increased odds of incident urgency UI (≥19.4% vs <14.1% of energy intake OR = 1.14; 95% CI 0.99, 1.30; p trend = .02), while CIs were wide and included 1.0 for calibrated protein intake. Other macronutrients were not associated with urgency UI and macronutrient intake was not associated with incident stress or mixed UI (p trend > .05 for all).
Conclusions
Among postmenopausal women, higher physical activity was associated with lower risk of incident urgency and mixed UI, but not stress UI, independent of baseline weight and weight change. Higher protein intake was associated with increased risk of urgency UI, but no associations were observed between other macronutrient and UI subtypes.
Keywords: Carbohydrates, Dietary fat, Dietary protein, Exercise, Macronutrients, Urology
Urinary incontinence (UI) affects nearly 20% of women over age 50 and, accounting for the aging population, is expected to affect more than 43 million American women over the next 30 years (1,2). UI accounts for more than $60 billion in annual direct costs in the United States (3). Among older women, UI is a geriatric syndrome associated with poor quality of life, falls, fractures, and functional decline (4–6). Older age and obesity are 2 established risk factors for UI (7–9). Weight loss, via surgery (10) or a multimodal lifestyle intervention (11–13), lowers UI risk and severity among overweight and obese women. These interventions induce weight loss by achieving a negative energy balance—the sum of energy expenditure and energy intake (14). However, it remains unknown if the primary determinants of energy balance, physical activity (energy expenditure) and macronutrient intake (energy intake), are associated with UI via pathways independent of weight changes. This is particularly important for preventing UI as women age because weight loss in older adults may be associated with increased morbidity and mortality (15,16).
The mechanism of weight loss leading to recovery or maintenance of continence among women remains unknown and may differ by UI subtype. Weight loss may influence UI through direct effects of lower adiposity on UI (17), greater physical activity and decreased sedentary time (12,13), preventing cardiovascular or metabolic risk factors (18), or changes in amount or relative contribution of macronutrients (19,20) (carbohydrate, protein, fat, and alcohol). Components of weight loss interventions may also influence risk of UI independent of weight loss. For example, greater physical activity may lead to increased physiologic reserve via pleotropic effects or directly strengthen pelvic floor muscles (6,21), changes in body composition such as decreased waist circumference (8), and lower dietary fat intake may lead to decreased circulating estrogen levels, which are associated with risk of UI (22). However, prior studies are limited by insufficient sample size to examine UI subtypes, lack of adjustment for important confounders, such as change in weight and total energy intake, and cross-sectional study design. It remains unknown and, therefore, we evaluated whether physical activity or macronutrient intake are independently associated with incident urgency, stress, or mixed UI among postmenopausal women.
We evaluated the association between the 2 major components of energy balance—physical activity and macronutrient intake—and incident urgency, stress, and mixed UI among postmenopausal women in the Women’s Health Initiative (WHI) Observational Study. We hypothesized that higher physical activity and lower dietary fat intake would be inversely associated with incident urgency, stress, and mixed UI after adjusting for baseline weight, weight change, and other confounding variables.
Method
Participants
WHI Observational Study is a prospective, multicenter cohort study of 93 676 postmenopausal women. Detailed methods have been previously published elsewhere (23). Briefly, participants were identified at 40 clinical centers across the United States, enrolled from 1993 to 1998, and followed for a median of 8 years. Women completed several self-administered questionnaires and WHI staff collected anthropometric measures at enrollment and throughout follow-up. Participants were excluded if at baseline they reported prevalent UI in the past 12 months (n = 64 061) or a history of stroke, multiple sclerosis, Alzheimer’s or Parkinson’s disease (n = 1189), did not complete the food frequency questionnaire (FFQ) or reported unrealistic caloric intake (defined as <600 or >5000 kcal/day; n = 1513), were missing weight (n = 1105) or other covariate data (n = 4126), or did not complete the UI questionnaire during follow-up (n = 2751) (Supplementary Figure 1).
Physical Activity Assessment
Physical activity was assessed at baseline with self-administered questionnaires. Participants were asked the frequency and duration they currently walked outside of their home or participated in low-intensity (eg, slow dancing, bowling, golf), moderate-intensity (eg, biking outdoors, using an exercise machine, calisthenics, easy swimming), and high-intensity (eg, aerobics, jogging, tennis, swimming laps) physical activities. Categories of frequency included rarely/never, 1–3 times per week, 4–6 times per week, and 7 or more times per week. Duration categories included less than 20 minutes, 20–39 minutes, 40–59 minutes, and 1 hour or more. Walking was further categorized into 4 speed categories which included less than 2 mph (casual), 2–3 mph (average), 3–4 mph (fast), and more than 4 mph (very fast). We then imputed the midpoint value for ranges of both frequency and duration and assigned an average metabolic equivalent task (MET) to each activity type, classifying each by intensity type (low, moderate, and high intensity), and for walking, by walking pace (average, fast, and very fast). We multiplied minutes * frequency * average MET and summed values across all activity types to generate a current total physical activity variable (MET-hours/week) consistent with prior WHI studies (24,25). For use in sensitivity analyses, we also calculate MET-hours/week separately for walking, low-intensity, moderate-intensity, and vigorous-intensity activity. Among a random sample of 536 women in this cohort who completed physical activity assessments approximately 10 weeks after baseline, the test–retest reliability (weighted ƙ) for individual physical activity variables ranged from 0.53 to 0.72 and intraclass correlation for total physical activity was 0.77 (26).
Dietary Assessment
Participants completed the self-administered WHI FFQ at baseline, which includes 19 adjustment questions (predominantly to refine fat intake estimates by asking about food preparation practices and added fats), 122 individual foods and food groups, and 4 summary questions. Individual food items and food groups were further quantified using questions on usual frequency and portion size. Total energy intake and macronutrient values were calculated using the Nutrition Data System for Research (version 2005; University of Minnesota, Minneapolis, MN). Macronutrient-specific energy intake was calculated by multiplying total grams per macronutrient by an energy density factor (9 for fat, 7 for alcohol, 4 for both protein and carbohydrates) and summed across all macronutrients to generate total energy intake. Percent total energy intake from each macronutrient was calculated by dividing the macronutrient-specific energy intake by total energy intake. Energy-adjusted correlation coefficients between the WHI FFQ and 4-day food records plus 4 24-hour dietary recalls are 0.58, 0.61, and 0.41 for percent energy from fat, carbohydrate, and protein, respectively (27). Since calibration studies have demonstrated that protein intake is overreported using FFQ compared to biomarker-calibrated estimates (28), calibrated protein intake was calculated by linear regression of log-transformed urinary nitrogen on log-FFQ estimates of protein intake in addition to body mass index, age, income, water intake, number of live births, self-reported health status, and duration since menopause. There are no established biomarkers for fat or carbohydrate intake, so we were only able to calculate calibrated protein intake and report all other macronutrient intakes as uncalibrated values.
UI Assessment
Women completed self-administered questionnaires that asked detailed questions on UI based on similar items used and validated in previous large epidemiologic studies (5,22,29) at baseline and after 3 years of follow-up. Prevalent UI was defined at baseline as answering yes to the question “Have you ever leaked even a very small amount of urine involuntarily and you couldn’t control it?” and answering “Less than once a month,” “More than once a month but less than once a week,” “One or more times a week but less than every day,” or “Daily” or more frequently in a subsequent question asking “How frequently does this occur?.” Incident UI was defined using the same questions and definition on the follow-up questionnaire among women without prevalent UI at baseline. In sensitivity analyses, “at least monthly UI” was defined as a response of “More than once a month but less than once per week” or more frequently. We further categorized UI subtype by response to the question “When do you usually leak urine?” Participants could select multiple responses and were defined as having urgency UI if they selected only “When I feel the need to urinate and can’t get to the toilet fast enough,” stress UI if they selected only “When I cough, laugh, sneeze, lift, stand up or exercise,” and mixed UI if they selected both or “Other.”
Data Analysis
We separately evaluated the association of physical activity and macronutrient intake with incident UI using multivariable logistic regression. Primary exposures included total physical activity (MET-hours/week), calibrated and uncalibrated protein intake (% energy intake), fat intake (% energy intake), carbohydrate intake (% energy intake), and alcohol intake (% energy intake). We categorized physical activity using cutoffs from prior WHI analyses (24) and examined macronutrients by quintile of intake. The basic model was adjusted for age, number of live births, number of pregnancies, and total energy intake based on causal knowledge from the prior literature (12). The full multivariable adjusted models included additional suspected confounders: age (continuous in years), number of live births (0, 1, 2–4, ≥5), number of pregnancies (0, 1, 2–4, ≥5), total energy intake (continuous in kcal/week), race/ethnicity (White, Black, Hispanic, or Other), income (<$10 000, $10 000–19 999, $20 000–34 999, $35 000–49 999, $50 000–$74 999, ≥$75 000), history of diabetes mellitus or heart failure (yes/no), self-reported health status (fair–poor, good, very good, excellent), time since menopause at baseline (continuous in years), ever use of hormone replacement therapy (yes/no), water intake (continuous in g/day), coffee intake (continuous in g/day), and alcohol intake (% of energy) assessed via baseline questionnaire. Weight and height were measured by study staff during the baseline and follow-up visit and used to calculate body mass index (kg/m2). Since weight is both a potential confounder and a potential mediator of the association between physical activity or macronutrient intake and UI subtypes (Supplementary Figure 2), models were further adjusted for baseline weight and change in weight (follow-up minus baseline) and results were reported separately. Additional covariates were considered based on a priori knowledge of the causal framework but not included in the final models because they were not significantly associated with UI in bivariate models: smoking status, diuretic use, and history of myocardial infarction, hypertension, or hyperlipidemia (p > .10 for all).
For models examining relationships with fat, protein, and alcohol intake, we evaluated the association between consuming higher amounts of each macronutrient and lower amounts of carbohydrate, the largest contributor to calories in the U.S. diet, while holding total calorie intake constant using the multivariate nutrient density method (30,31). Therefore, we included percent energy from calibrated and uncalibrated protein, alcohol, and fat in our multivariate models along with total calories and other covariates (described above) and omitted percent energy from carbohydrates. The effect estimate for the macronutrient of interest can be interpreted as the effect of increasing calories from that macronutrient and decreasing calories from carbohydrate by the same amount. For models examining relationships with carbohydrate intake, we evaluated the association between consuming higher amounts of carbohydrate and lower amounts of fat by omitting percent energy intake from fat.
We conducted several sensitivity analyses using an alternate case definition for incident UI (“at least monthly”), separate independent variables for walking and various levels of physical activity intensity (low, moderate, and vigorous intensity), further adjusting for waist circumference in models adjusted for weight, and excluding women who had undergone a hysterectomy because this surgery could cause UI through mechanisms unrelated to our relationship of interest (32). p Trend was calculated across categories of exposures using a linear trend test. For calibrated protein intake, the standard errors calculation of CIs and p trend were estimated using a bootstrap resampling procedure with 10 000 resamples. p < .05 was considered statistically significant. All analyses were performed using R version 3.4.4 (R Foundation for Statistical Computing, Vienna, Austria).
Results
Table 1 displays the baseline characteristics of 19 741 women with and without incident urgency, stress, or mixed UI during follow-up. Women who developed any UI subtype were somewhat less likely to report “excellent” health (17%–21% vs 24%) compared to women who did not develop UI. Some differences were also observed across UI subtypes; women with incident urgency or mixed UI were more likely to report a greater burden of comorbidities and use of medications and women with incident stress UI were more likely to ever use hormone replacement therapy. Calibrated energy intake was similar between women with and without incident UI.
Table 1.
Baseline Characteristics of 19 741 Women in the Analytic Cohort, by Incontinence Status at Year 3
| None | Incident Urgency UI | Incident Stress UI | Incident Mixed UI | |
|---|---|---|---|---|
| Age at screening, years, n (%) | ||||
| 50–54 | 2186 (16) | 315 (11) | 387 (16) | 85 (16) |
| 55–59 | 2839 (20) | 479 (17) | 524 (22) | 90 (16) |
| 60–69 | 3209 (23) | 569 (21) | 574 (24) | 136 (25) |
| 70–79 | 2980 (21) | 648 (24) | 491 (20) | 112 (20) |
| Race/ethnicity, n (%) | ||||
| Black | 1498 (11) | 284 (10) | 133 (5.5) | 49 (8.9) |
| Hispanic | 469 (3.3) | 75 (2.7) | 78 (3.2) | 21 (3.8) |
| White | 11 277 (80) | 2249 (82) | 2038 (85) | 452 (82) |
| Other | 784 (5.6) | 153 (5.5) | 153 (6.4) | 28 (5.1) |
| Current smoking, n (%) | 841 (6.0) | 161 (5.8) | 135 (5.6) | 38 (6.9) |
| Family income, n (%) | ||||
| <$10 000 | 541 (3.9) | 113 (4.1) | 67 (2.8) | 33 (6.0) |
| $10 000–$19 999 | 1469 (11) | 326 (12) | 214 (8.9) | 60 (11) |
| $20 000–$34 999 | 3164 (23) | 684 (25) | 560 (23) | 138 (25) |
| $35 000–$49 999 | 2773 (20) | 582 (21) | 502 (21) | 110 (20) |
| $50 000–$74 999 | 2926 (21) | 528 (19) | 513 (21) | 104 (19) |
| $75 000+ | 3155 (23) | 528 (19) | 546 (23) | 105 (19) |
| Self-reported health status, n (%) | ||||
| Excellent | 3364 (24) | 505 (18) | 514 (21) | 94 (17) |
| Very good | 6064 (43) | 1186 (43) | 1054 (44) | 229 (42) |
| Good | 3768 (27) | 865 (31) | 690 (29) | 174 (32) |
| Fair or poor | 832 (5.9) | 205 (7.4) | 144 (6.0) | 53 (9.6) |
| Years since menopause, mean (SD) | 14.3 (9.1) | 15.9 (9.2) | 13.9 (8.8) | 14.7 (8.9) |
| Number of live births, n (%) | ||||
| 0 | 2374 (17) | 384 (14) | 360 (15) | 92 (17) |
| 1 | 1403 (10) | 272 (10) | 245 (10) | 46 (8.4) |
| 2–4 | 8833 (63) | 1746 (63) | 1550 (65) | 353 (64) |
| 5+ | 1418 (10) | 359 (13) | 247 (10) | 59 (11) |
| Number of pregnancies | ||||
| 0 | 1847 (13) | 299 (11) | 291 (12) | 73 (13) |
| 1 | 1165 (8.3) | 207 (7.5) | 211 (8.8) | 39 (7.1) |
| 2–4 | 8287 (59) | 1630 (59) | 1426 (59) | 316 (58) |
| 5+ | 2714 (19) | 622 (23) | 473 (20) | 122 (22) |
| Comorbidities, n (%) | ||||
| Hypertension | 3898 (28) | 909 (33) | 689 (29) | 172 (31) |
| Diabetes mellitus | 546 (3.9) | 161 (5.8) | 85 (3.5) | 34 (6.2) |
| Myocardial infarction | 231 (1.6) | 63 (2.3) | 40 (1.7) | 15 (2.7) |
| Heart failure | 88 (0.6) | 27 (1.0) | 13 (0.5) | 7 (1.3) |
| Medication use, n (%) | ||||
| Ever use hormone replacement therapy | 7818 (56) | 1613 (58) | 1532 (64) | 314 (57) |
| Treatment for hypertension (excluding diuretics) | 2930 (21) | 670 (24) | 517 (22) | 129 (24) |
| Treatment for high cholesterol | 1866 (13) | 421 (15) | 352 (15) | 73 (13) |
| Diuretics (thiazides and loop) | 652 (4.6) | 140 (5.1) | 90 (3.7) | 32 (5.8) |
| Treatment for diabetes | 394 (2.8) | 112 (4.1) | 51 (2.1) | 27 (4.9) |
| Water intake, g/day, mean (SD) | 1544 (635) | 1575 (620) | 1570 (601) | 1581 (618) |
| Caffeine intake, mg/day, mean (SD) | 156 (129) | 163 (129) | 160 (125) | 166 (138) |
| Energy intake, kcal/day, mean (SD) | ||||
| Raw | 1499 (573) | 1539 (578) | 1523 (581) | 1558 (539) |
| Calibrated* | 2065 (228) | 2064 (234) | 2079 (220) | 2081 (244) |
| Height, cm, mean (SD) | 162 (6.6) | 162 (6.5) | 162 (6.6) | 162 (6.1) |
| Weight, kg, mean (SD) | 68.3 (14) | 70.3 (15) | 68.6 (14) | 71.5 (16) |
| BMI, kg/m2, n (%) | ||||
| <25 | 6922 (49) | 1206 (44) | 1164 (49) | 236 (43) |
| 25 to <30 | 2485 (18) | 625 (23) | 448 (19) | 147 (27) |
| ≥30 | 4621 (33) | 930 (34) | 790 (33) | 167 (30) |
| Waist circumference, cm, mean (SD) | 81.7 (12) | 83.9 (13) | 82.1 (12) | 85.1 (14) |
Notes: BMI = body mass index; kcal = kilocalories; n = sample size; SD = standard deviation; UI = urinary incontinence.
*Calibrated energy intake was modeled using BMI, age, income, water intake, number of live births, self-reported health status, and duration since menopause.
The odds of developing urgency, stress, and mixed UI within categories of total physical activity (MET-hours/week) are displayed in Table 2. In minimally adjusted (Model 1) and multivariable (Model 2) models, higher levels of physical activity were inversely associated with odds of incident urgency and mixed UI, but not stress UI (p trend = .13). Women who reported ≥30 versus <0.1 MET-hours/week of total physical activity had 21% and 38% lower odds of incident urgency and mixed UI after 3 years of follow-up, respectively (urgency UI: OR = 0.79; 95% CI 0.67, 0.94; p trend = .03; mixed UI: OR = 0.62; 95 % CI 0.43, 0.87; p trend = .04). After further adjustment for baseline weight and weight change, women reporting ≥30 MET-hours/week were still less likely to develop urgency and mixed UI compared to those reporting <0.1 MET-hours/week; however, these associations were slightly attenuated and the linear trends were no longer statistically significant (Model 3; p trend = .15 for urgency UI and 0.16 for mixed UI). When incident UI was defined as “at least monthly” UI episodes (Supplementary Table 1), associations between total physical activity and incident urgency and mixed UI were further attenuated and CIs were wider (urgency UI: p trend = .04; mixed UI: p trend = .15), whereas greater total physical activity was inversely associated with incident stress UI (p trend = .02).
Table 2.
Odds of Incident Urinary Incontinence After 3 Years of Follow-up Among 19 741 Postmenopausal Women, by Incontinence Subtype and Category of Total Physical Activity*
| Category of Total Physical Activity, MET-hours/week | |||||||
|---|---|---|---|---|---|---|---|
| <0.1 (n = 2202) | 0.1–4.9 (n = 3430) | 5–9.9 (n = 3273) | 10–19.9 (n = 5167) | 20–29.9 (n = 2808) | ≥30 (n = 2861) | p Trend | |
| Urgency urinary incontinence | |||||||
| Number of cases | 345 | 526 | 443 | 721 | 395 | 331 | |
| Model 1 | Ref. | 0.96 (0.83, 1.12) | 0.84 (0.72, 0.98) | 0.87 (0.76, 1.00) | 0.90 (0.77, 1.05) | 0.72 (0.61, 0.85) | <.001 |
| Model 2 | Ref. | 0.97 (0.83, 1.13) | 0.87 (0.74, 1.01) | 0.91 (0.79, 1.05) | 0.96 (0.82, 1.13) | 0.79 (0.67, 0.94) | .03 |
| Model 3 | Ref. | 0.99 (0.85, 1.16) | 0.87 (0.74, 1.03) | 0.94 (0.81, 1.09) | 1.01 (0.85, 1.19) | 0.84 (0.70, 1.00) | .15 |
| Stress urinary incontinence | |||||||
| Number of cases | 290 | 382 | 420 | 647 | 336 | 327 | |
| Model 1 | Ref. | 0.83 (0.71, 0.98) | 0.98 (0.83, 1.15) | 0.95 (0.82, 1.10) | 0.90 (0.76, 1.06) | 0.85 (0.72, 1.00) | .30 |
| Model 2 | Ref. | 0.82 (0.69, 0.96) | 0.94 (0.80, 1.10) | 0.90 (0.78, 1.05) | 0.85 (0.71, 1.01) | 0.82 (0.68, 0.97) | .13 |
| Model 3 | Ref. | 0.86 (0.73, 1.03) | 1.00 (0.85, 1.19) | 0.97 (0.82, 1.13) | 0.89 (0.74, 1.07) | 0.91 (0.76, 1.09) | .55 |
| Mixed urinary incontinence | |||||||
| Number of cases | 88 | 96 | 88 | 147 | 70 | 61 | |
| Model 1 | Ref. | 0.69 (0.51, 0.93) | 0.66 (0.49, 0.90) | 0.70 (0.54, 0.92) | 0.62 (0.45, 0.85) | 0.53 (0.38, 0.73) | <.001 |
| Model 2 | Ref. | 0.70 (0.52, 0.94) | 0.70 (0.52, 0.95) | 0.77 (0.59, 1.02) | 0.70 (0.50, 0.97) | 0.62 (0.43, 0.87) | .04 |
| Model 3 | Ref. | 0.68 (0.50, 0.93) | 0.71 (0.51, 0.97) | 0.82 (0.61, 1.09) | 0.72 (0.51, 1.01) | 0.66 (0.46, 0.95) | .16 |
Notes: MET = metabolic equivalent task.
*Odds ratios were calculated using a multivariate logistic regression model and p trends were calculated using a linear trend test. Model 1 adjusted for age, number of live births, number of pregnancies, and total calories. Model 2 adjusted for Model 1 + race/ethnicity, income, diabetes mellitus, heart failure, self-reported health status, years since menopause at baseline, hormone replacement therapy, coffee intake, water intake, and alcohol intake (percentage of energy intake). Model 3 adjusted for Model 2 + height and weight at baseline and change in weight (follow-up minus baseline).
The strength of observed associations varied by walking and different intensity levels of physical activity (low, moderate, and vigorous intensity) and the strongest inverse associations were observed between walking and incident urgency and mixed UI as well as moderate-intensity physical activity and incident urgency UI (Table 3). Women who reported ≥10.4 versus 0 MET-hours/week of walking had 13% and 36% lower odds of incident urgency and mixed UI, respectively, independent of baseline weight and weight change (urgency UI: OR = 0.87; 95% CI 0.76, 0.99; p trend = .08; mixed UI: OR = 0.64; 95 % CI 0.48, 0.84; p trend = .001). Low-intensity physical activity was not associated with any UI subtypes and moderate-intensity physical activity was inversely associated with risk of incident urgency UI (p trend = .01), but not stress or mixed UI. There was no linear trend observed for vigorous-intensity physical activity and any UI subtype (p trend > .05 for all); however, women who reported the most vigorous-intensity activity (≥23.3 vs 0 MET-hours/week) were 19% less likely to develop stress UI (highest vs lowest quintile OR = 0.81; 95% CI 0.67, 0.97).
Table 3.
Odds of Incident Urinary Incontinence After 3 Years of Follow-up Among 19 741 Postmenopausal Women, by Incontinence Subtype and Category of Walking, Light, Moderate, and Vigorous Physical Activity*
| Quintile of Walking MET-hours/week | ||||||
|---|---|---|---|---|---|---|
| 0 (n = 5227) | 0.1–<3.1 (n = 3398) | 3.1–<5.3 (n = 3493) | 5.3–<10.4 (n = 3779) | ≥10.4 (n = 3844) | p Trend | |
| Urgency urinary incontinence | ||||||
| Number of cases | 776 | 510 | 478 | 526 | 471 | |
| Model 1 | Ref. | 1.00 (0.88, 1.12) | 0.91 (0.80, 1.03) | 0.93 (0.82, 1.05) | 0.81 (0.71, 0.91) | <.001 |
| Model 2 | Ref. | 0.99 (0.88, 1.12) | 0.93 (0.82, 1.05) | 0.95 (0.84, 1.08) | 0.84 (0.74, 0.95) | .01 |
| Model 3 | Ref. | 1.01 (0.89, 1.15) | 0.93 (0.81, 1.06) | 1.01 (0.89, 1.14) | 0.87 (0.76, 0.99) | .08 |
| Stress urinary incontinence | ||||||
| Number of cases | 638 | 422 | 425 | 490 | 427 | |
| Model 1 | Ref. | 1.03 (0.90, 1.17) | .99 (0.87, 1.13) | 1.07 (0.95, 1.22) | 0.90 (0.79, 1.02) | .33 |
| Model 2 | Ref. | 1.01 (0.89, 1.16) | 0.97 (0.85, 1.11) | 1.05 (0.92, 1.19) | 0.89 (0.78, 1.02) | .25 |
| Model 3 | Ref. | 1.07 (0.93, 1.22) | 1.03 (0.90, 1.18) | 1.08 (0.95, 1.24) | 0.93 (0.81, 1.07) | .59 |
| Mixed urinary incontinence | ||||||
| Number of cases | 192 | 93 | 86 | 94 | 85 | |
| Model 1 | Ref. | 0.73 (0.57, 0.94) | 0.66 (0.51, 0.85) | 0.67 (0.52, 0.85) | 0.59 (0.46, 0.77) | <.001 |
| Model 2 | Ref. | 0.73 (0.56, 0.93) | 0.68 (0.52, 0.88) | 0.70 (0.54, 0.89) | 0.63 (0.48, 0.81) | <.001 |
| Model 3 | Ref. | 0.73 (0.56, 0.94) | 0.70 (0.53, 0.91) | 0.72 (0.55, 0.93) | 0.64 (0.48, 0.84) | .001 |
| Quintile of Low-Intensity Physical Activity MET-hours/week | ||||||
| 0 (n = 14 152) | 0.1–<3.0 (n = 1771) | 3.0–<4.5 (n = 1605) | 4.5–<7.5 (n = 1119) | ≥7.5 (n = 1094) | p Trend | |
| Urgency urinary incontinence | ||||||
| Number of cases | 2001 | 245 | 210 | 159 | 146 | |
| Model 1 | Ref. | 0.99 (0.86, 1.14) | 0.92 (0.78, 1.07) | 0.98 (0.82, 1.16) | 0.91 (0.75, 1.08) | .22 |
| Model 2 | Ref. | 1.00 (0.86, 1.15) | 0.93 (0.79, 1.08) | 0.99 (0.83, 1.18) | 0.94 (0.78, 1.12) | .40 |
| Model 3 | Ref. | 1.04 (0.89, 1.20) | 0.95 (0.81, 1.11) | 1.06 (0.88, 1.26) | 0.96 (0.79, 1.15) | .85 |
| Stress urinary incontinence | ||||||
| Number of cases | 1677 | 228 | 219 | 152 | 126 | |
| Model 1 | Ref. | 1.09 (0.94, 1.26) | 1.17 (1.01, 1.36) | 1.18 (0.98, 1.40) | 0.98 (0.80, 1.18) | .12 |
| Model 2 | Ref. | 1.09 (0.94, 1.26) | 1.14 (0.98, 1.33) | 1.14 (0.95, 1.36) | 0.95 (0.78, 1.15) | .30 |
| Model 3 | Ref. | 1.12 (0.96, 1.30) | 1.14 (0.97, 1.33) | 1.16 (0.96, 1.39) | 0.99 (0.81, 1.20) | .17 |
| Mixed urinary incontinence | ||||||
| Number of cases | 399 | 51 | 35 | 33 | 32 | |
| Model 1 | Ref. | 1.02 (0.75, 1.36) | 0.77 (0.53, 1.07) | 1.04 (0.71, 1.46) | 1.02 (0.69, 1.44) | .74 |
| Model 2 | Ref. | 1.03 (0.76, 1.38) | 0.79 (0.55, 1.11) | 1.08 (0.74, 1.52) | 1.07 (0.73, 1.52) | .97 |
| Model 3 | Ref. | 1.01 (0.73, 1.37) | 0.72 (0.49, 1.04) | 1.14 (0.77, 1.62) | 1.18 (0.80, 1.67) | .71 |
| Quintile of Moderate-Intensity Physical Activity MET-hours/week | ||||||
| 0 (n = 9783) | 0.1–<3.0 (n = 2928) | 3.0–<6.8 (n = 2375) | 6.8–<10.5 (n = 2179) | ≥10.5 (n = 2476) | p Trend | |
| Urgency urinary incontinence | ||||||
| Number of cases | 1440 | 406 | 310 | 289 | 316 | |
| Model 1 | Ref. | 0.96 (0.85, 1.08) | 0.88 (0.77, 1.00) | 0.91 (0.79, 1.04) | 0.84 (0.74, 0.96) | <.001 |
| Model 2 | Ref. | 0.96 (0.85, 1.09) | 0.90 (0.79, 1.03) | 0.93 (0.81, 1.07) | 0.88 (0.77, 1.00) | .03 |
| Model 3 | Ref. | 0.96 (0.85, 1.08) | 0.88 (0.77, 1.01) | 0.91 (0.78, 1.04) | 0.86 (0.75, 0.98) | .01 |
| Stress urinary incontinence | ||||||
| Number of cases | 1146 | 365 | 324 | 275 | 292 | |
| Model 1 | Ref. | 1.05 (0.93, 1.20) | 1.19 (1.04, 1.35) | 1.08 (0.93, 1.24) | 1.01 (0.88, 1.16) | .30 |
| Model 2 | Ref. | 1.05 (0.93, 1.19) | 1.16 (1.02, 1.33) | 1.05 (0.91, 1.21) | 0.99 (0.86, 1.14) | .56 |
| Model 3 | Ref. | 1.07 (0.93, 1.21) | 1.18 (1.02, 1.35) | 1.06 (0.92, 1.23) | 1.05 (0.91, 1.21) | .20 |
| Mixed urinary incontinence | ||||||
| Number of cases | 284 | 73 | 62 | 66 | 65 | |
| Model 1 | Ref. | 0.86 (0.66, 1.11) | 0.90 (0.67, 1.18) | 1.04 (0.79, 1.36) | 0.90 (0.68, 1.18) | .62 |
| Model 2 | Ref. | 0.88 (0.67, 1.14) | 0.95 (0.71, 1.25) | 1.11 (0.84, 1.46) | 0.97 (0.73, 1.28) | .85 |
| Model 3 | Ref. | 0.86 (0.65, 1.12) | 0.92 (0.68, 1.23) | 1.12 (0.84, 1.48) | 0.99 (0.74, 1.31) | .78 |
| Quintile of Vigorous-Intensity Physical Activity MET-hours/week | ||||||
| 0 (n = 13 821) | 0.1–<8.2 (n = 1800) | 8.2–<14.0 (n = 1319) | 14.0–<23.3 (n = 1483) | ≥23.3 (n = 1318) | p Trend | |
| Urgency urinary incontinence | ||||||
| Number of cases | 2005 | 237 | 172 | 190 | 157 | |
| Model 1 | Ref. | 0.93 (0.80, 1.07) | 0.93 (0.78, 1.09) | 0.91 (0.78, 1.07) | 0.83 (0.70, 0.99) | .02 |
| Model 2 | Ref. | 0.96 (0.83, 1.11) | 0.97 (0.81, 1.14) | 0.97 (0.82, 1.14) | 0.90 (0.75, 1.07) | .25 |
| Model 3 | Ref. | 0.97 (0.84, 1.13) | 0.95 (0.80, 1.13) | 1.01 (0.85, 1.19) | 0.95 (0.79, 1.14) | .62 |
| Stress urinary incontinence | ||||||
| Number of cases | 1697 | 218 | 170 | 182 | 135 | |
| Model 1 | Ref. | 0.97 (0.83, 1.12) | 1.04 (0.87, 1.22) | 0.98 (0.83, 1.16) | 0.80 (0.66, 0.96) | .09 |
| Model 2 | Ref. | 0.98 (0.84, 1.14) | 1.04 (0.87, 1.23) | 0.98 (0.83, 1.16) | 0.81 (0.67, 0.97) | .12 |
| Model 3 | Ref. | 1.01 (0.86, 1.18) | 1.01 (0.84, 1.20) | 0.99 (0.83, 1.17) | 0.86 (0.71, 1.04) | .27 |
| Mixed urinary incontinence | ||||||
| Number of cases | 411 | 39 | 34 | 31 | 35 | |
| Model 1 | Ref. | 0.73 (0.51, 1.00) | 0.87 (0.60, 1.22) | 0.70 (0.48, 1.00) | 0.90 (0.62, 1.26) | .08 |
| Model 2 | Ref. | 0.78 (0.55, 1.07) | 0.95 (0.65, 1.34) | 0.78 (0.53, 1.12) | 1.04 (0.71, 1.46) | .47 |
| Model 3 | Ref. | 0.83 (0.58, 1.16) | 0.95 (0.64, 1.35) | 0.84 (0.56, 1.21) | 1.04 (0.70, 1.49) | .61 |
Notes: MET = metabolic equivalent task.
*Participants categorized into 0 MET-hours/week and rough quartiles among those with >0 MET-hours/week for each activity each. Odds ratios were calculated using a multivariate logistic regression model and p trends were calculated using a linear trend test. Model 1 adjusted for age, number of live births, number of pregnancies, and total calories. Model 2 adjusted for Model 1 + race/ethnicity, income, diabetes mellitus, heart failure, self-reported health status, years since menopause at baseline, hormone replacement therapy, coffee intake, water intake, and alcohol intake (percentage of energy intake). Model 3 adjusted for Model 2 + height and weight at baseline and change in weight (follow-up minus baseline).
Table 4 displays the odds of developing urgency UI by quintile of macronutrient intake—specifically percent energy intake from protein (calibrated and uncalibrated), fat, carbohydrate, and alcohol. Replacing energy from carbohydrate with uncalibrated protein was associated with increased odds of incident urgency UI (highest vs lowest quintile of % energy from uncalibrated protein OR = 1.16; 95% CI 1.01, 1.32; p trend = .01), and this association was independent of baseline weight and weight change (p trend = .02). Associations with percent energy intake from calibrated protein intake were in the same direction as uncalibrated protein intake, but the ORs were larger and the CIs were wider (highest vs lowest quintile OR = 1.50; 95% CI 0.91, 2.47; p trend = .13). In sensitivity analyses, uncalibrated protein intake was no longer associated with “at least monthly” urgency UI (p trend = .18) (Supplementary Table 2). Percent energy intake from fat, carbohydrate, and alcohol intake were not associated risk of urgency UI regardless of adjustment for weight (p trend < .05 for all).
Table 4.
Odds of Incident Urgency Urinary Incontinence After 3 Years of Follow-up Among 19 741 Postmenopausal Women, by Quintile of Macronutrient Intake*
| Quintile‡ of Macronutrient Intake | ||||||
|---|---|---|---|---|---|---|
| Calibrated Protein†, % Energy Intake | <12.7 | 12.7–<13.8 | 13.8–<14.7 | 14.7–<15.8 | ≥15.8 | p Trend |
| Number of cases | 537 | 562 | 554 | 547 | 561 | |
| Model 1 | Ref. | 1.06 (0.87, 1.30) | 1.07 (0.86, 1.34) | 1.09 (0.85, 1.39) | 1.18 (0.89, 1.58) | .29 |
| Model 2 | Ref. | 1.19 (0.94, 1.52) | 1.26 (0.93, 1.70) | 1.32 (0.92, 1.90) | 1.50 (0.91, 2.47) | .13 |
| Model 3 | Ref. | 1.13 (0.90, 1.42) | 1.19 (0.91, 1.56) | 1.24 (0.91, 1.71) | 1.35 (0.89, 2.05) | .17 |
| Uncalibrated Protein, % Energy Intake | <14.1 | 14.1–<15.9 | 15.9–<17.5 | 17.5–<19.4 | ≥19.4 | |
| Number of cases | 524 | 513 | 567 | 575 | 582 | |
| Model 1 | Ref. | 0.97 (0.85, 1.10) | 1.09 (0.96, 1.24) | 1.12 (0.98, 1.27) | 1.15 (1.01, 1.31) | .01 |
| Model 2 | Ref. | 0.97 (0.85, 1.11) | 1.09 (0.96, 1.24) | 1.12 (0.98, 1.27) | 1.16 (1.01, 1.32) | .01 |
| Model 3 | Ref. | 0.99 (0.86, 1.14) | 1.10 (0.96, 1.26) | 1.13 (0.98, 1.29) | 1.14 (0.99, 1.30) | .02 |
| Fat, % Energy Intake | <22.3 | 22.3–<26.8 | 26.8–<31.0 | 31.0–<36.9 | ≥36.9 | |
| Number of cases | 483 | 546 | 568 | 583 | 581 | |
| Model 1 | Ref. | 1.14 (1.00, 1.30) | 1.19 (1.04, 1.35) | 1.20 (1.05, 1.36) | 1.20 (1.05, 1.37) | .01 |
| Model 2 | Ref. | 1.09 (0.96, 1.25) | 1.12 (0.97, 1.28) | 1.10 (0.95, 1.27) | 1.05 (0.89, 1.24) | .59 |
| Model 3 | Ref. | 1.06 (0.92, 1.22) | 1.09 (0.95, 1.26) | 1.06 (0.91, 1.23) | 1.02 (0.86, 1.21) | .82 |
| Carbohydrates, % Energy Intake | <45.0 | 45.0–<51.0 | 51.0–<55.9 | 55.9–<61.4 | ≥61.4 | |
| Number of cases | 597 | 581 | 565 | 515 | 503 | |
| Model 1 | Ref. | 0.96 (0.85, 1.09) | 0.95 (0.83, 1.07) | 0.86 (0.76, 0.98) | 0.87 (0.76, 0.99) | .01 |
| Model 2 | Ref. | 0.99 (0.87, 1.12) | 0.99 (0.87, 1.13) | 0.92 (0.80, 1.05) | 0.95 (0.82, 1.11) | .32 |
| Model 3 | Ref. | 0.96 (0.85, 1.10) | 0.99 (0.86, 1.13) | 0.91 (0.79, 1.06) | 0.96 (0.82, 1.12) | .43 |
| Alcohol, % Energy Intake | 0 | 0.1–<0.7 | 0.7–<2.4 | 2.4–<6.1 | ≥6.1 | |
| Number of cases | 1157 | 412 | 388 | 394 | 410 | |
| Model 1 | Ref. | 0.92 (0.82, 1.04) | 0.90 (0.80, 1.02) | 0.91 (0.80, 1.02) | 0.97 (0.86, 1.10) | .26 |
| Model 2 | Ref. | 0.94 (0.83, 1.07) | 0.93 (0.81, 1.05) | 0.95 (0.83, 1.07) | 1.03 (0.90, 1.17) | .92 |
| Model 3 | Ref. | 0.92 (0.81, 1.05) | 0.92 (0.81, 1.05) | 0.95 (0.83, 1.09) | 1.02 (0.89, 1.17) | .96 |
Notes: BMI = body mass index.
*Odds ratios were calculated using a multivariable logistic regression model and p trends were calculated using a linear trend test. Model 1 adjusted for age, number of live births, number of pregnancies, and total calories. Model 2 adjusted for Model 1 + race/ethnicity, income, diabetes mellitus, heart failure, self-reported health status, years since menopause at baseline, hormone replacement therapy, coffee intake, and water intake. Model 3 adjusted for Model 2 + height and weight at baseline and change in weight (follow-up minus baseline). Fat and carbohydrate models were further adjusted for uncalibrated protein and alcohol intake (percentage of energy intake) and both calibrated and uncalibrated protein models were further adjusted for fat and alcohol intake (percentage of energy intake). Alcohol models were further adjusted for fat and uncalibrated protein intake (percentage of energy intake).
†Calibrated protein intake was modeled using BMI, age, income, water intake, number of live births, self-reported health status, and duration since menopause.
‡Except for alcohol intake which was categorized as 0 (n = 7906), 0.1–<0.7 (n = 2959), 0.7–<2.4 (n = 2959), 2.4–<6.1 (n = 2959), ≥6.1 (n = 2958).
Macronutrient intake was not significantly associated with odds of incident stress or mixed UI (Supplementary Tables 3 and 4). These results did not change after further adjustment for baseline weight and weight change or when considering “at least monthly” stress or mixed UI (Supplementary Tables 5 and 6). Overall results were unchanged when women with a history of hysterectomy were excluded (n = 6975) or after further adjustment for waist circumference (data available upon request).
Discussion
In this large, prospective, multicenter cohort study of postmenopausal women, physical activity and macronutrient intake were both associated with incident UI, independent of total energy intake, baseline weight, and weight change. Specifically, higher total physical activity was associated with lower odds of urgency and mixed UI, but not stress UI. The association between physical activity and UI subtypes was independent of baseline weight and weight change, although associations were consistently attenuated after adjusting for weight and varied by UI severity and physical activity intensity. Women who reported the equivalent of 3.5 hours/week of walking at an average pace or 2.5 hours/week of walking at a fast or very fast pace had the lowest risk of both urgency and mixed UI.
In this study, higher uncalibrated protein intake was associated with increased risk of incident urgency UI independent of baseline weight and weight change. However, this association was not robust in sensitivity analyses and CIs were wider with the more accurate measure of biomarker-calibrated protein intake. Fat, carbohydrate, and alcohol intake were not associated with risk of urgency UI and macronutrient intake overall was not associated with incident stress or mixed UI.
These results are consistent with prior literature demonstrating an inverse association between physical activity and risk of UI in women. In a large international study, women who were more physically active reported less prevalent UI, particularly urgency and mixed UI subtypes (33). Prospective cohort studies among both middle-aged (12) and predominantly postmenopausal women (13) have demonstrated a lower risk of incident UI with higher levels of physical activity; however, studies with smaller sample sizes have not observed the same association (18,34). Among women aged 54–79 years old enrolled in the Nurses’ Health Study, those in the highest quintile of total physical activity were less likely to develop incident UI compared to women in the lowest quintile (>28.6 vs ≤6.2 MET-hours/week OR = 0.81; 95% CI 0.71, 0.93; p trend < .01) (13). Similar to our study, this association was strongest for MET-hours/week spent walking (highest vs lowest quintile OR = 0.74; 95% CI 0.64, 0.88; p trend < .01). However, associations were only statistically significant for stress UI when examined separately by UI subtype. We observed a similar inverse association between total physical activity and incident “at least monthly” stress UI in sensitivity analyses, but did not observe this association in our primary models using a less severe UI definition, suggesting that the relationship between physical activity and stress UI may vary by stress UI severity. Conversely, our study observed more consistent inverse associations between total physical activity and incident urgency and mixed UI suggesting that the relationship between physical activity and UI may depend on both UI subtype and severity. Overall, our findings provide new evidence that total physical activity, and particularly walking, is inversely associated with urgency and mixed UI, independent of changes in weight.
The association between macronutrient intake and UI has been less well studied. A systematic review of lifestyle factors and lower urinary tract symptoms reported that the evidence grade was low for the majority of studies examining associations between diet and UI (70% level 3 or 4 evidence) (35). This includes 2 cross-sectional analyses that reported conflicting associations between dietary fat intake and prevalent UI, although both studies had methodologic limitations including lack of adjustment for weight and/or total energy intake (20,36). Among women aged 40–80 years old in the United Kingdom, those who reported higher carbohydrate intake were less likely to develop stress UI (highest vs lowest quintile OR = 0.73; 95% 0.49, 1.08; p trend = .05) and higher fat intake was associated with increased odds of incident stress UI (highest vs lowest quintile OR = 2.02; 95% CI 1.33, 3.05; p trend = .02) (19). However, macronutrient intake was not associated with overactive bladder, including urgency UI, in the same United Kingdom cohort (37). In our study, macronutrient intake was not associated with stress or mixed UI and only uncalibrated protein intake was positively associated with urgency UI. Taken together with the prior literature, our study provides preliminary evidence that a low-carbohydrate, high-protein diet may increase risk of urgency UI in postmenopausal women, independent of caloric intake and changes in weight.
Our findings suggest that the determinants of energy balance (14)—physical activity, macronutrient energy intake, and metabolic efficiency (not measured in our study but largely driven by body size (38))—are associated with UI, independent from the direct effect of weight. Several mechanisms for the association of physical activity and macronutrient intake with UI have been proposed. Increased physical activity may lead to lower risk of UI via increased generalized strength and stronger pelvic floor muscles (21), increased vaginal resting pressure (39), increased lower extremity strength and improved conditioning (6), decreased cerebral white matter intensities (40,41), and increased autonomic control (42,43). Physical activity could also lower risk of UI via decreased psychological stress, anxiety, and depression (44), which are established risk factors of UI, particularly urgency UI (45). The potential mechanisms of an association between macronutrient intake and UI are less well studied and remain exploratory. High-protein diets can theoretically cause urea-induced osmotic diuresis (46,47), which would increase the volume of urine in the bladder that could potentially leak. Although fat intake was not associated with UI in our study, high-fat diets may cause increased circulating estrogen levels, which are associated with increased risk of UI (eg, when increased via exogenous hormone replacement therapy in the original WHI trial) (22). Macronutrient intake also affects systemic inflammatory markers (48) and autonomic nervous system activity (49), although it remains unknown if these pathways affect local bladder inflammation or neuromodulation. Further studies are needed to elucidate the biological mechanisms and complex interactions of weight, physical activity, and macronutrient intake with respect to distinct UI subtypes.
This study has limitations. The observational study design limits our ability to control for confounding by unmeasured covariates, although demographic, clinical, and anthropometric measures were extensively collected in the WHI Observational Study. Self-reported UI frequency and subtypes are well-established and clinically relevant outcomes; however, we did not have additional quantitative measurements of amount of urine leaked or pad usage to supplement the self-reported questionnaires. As demonstrated in our comprehensive sensitivity analyses, associations appear to vary depending on UI severity; however, we did not have a sufficient sample size to examine associations with at least weekly UI. To select an analytic sample of women who are eligible to develop incident UI, we excluded a large number of women with any prevalent UI at baseline, which is a stricter inclusion criteria than prior studies and may have limited our ability to detect small effect sizes in subgroup analyses. Similarly, measurement error of self-reported physical activity and macronutrient intake could lead to imprecise effect estimates and potentially mask true associations. We report associations with each macronutrient class, but this does not eliminate the possibility that macronutrient subtypes may be positively or inversely associated with UI subtypes (eg, plant vs animal protein, saturated vs polyunsaturated vs monounsaturated fats, etc.). Self-reported physical activity METs are also poorly correlated with accelerometer-measured METs, particularly for low-intensity activities, which likely represent the majority of physical activity for women who are already at higher risk of UI due to older age or greater body mass index (50). Despite these limitations, this study provides high-quality observational data supporting the hypothesis that physical activity and macronutrient intake are associated with UI subtypes among postmenopausal women.
Conclusions
In conclusion, higher physical activity, particularly walking, is associated with lower risk of urgency and mixed UI. These associations were observed independent of changes in weight and may represent novel mechanisms of the effect of weight loss interventions on UI among postmenopausal women. Higher uncalibrated protein is associated with increased risk of incident urgency UI, but not other UI subtypes. Low-carbohydrate and high-protein dietary intervention studies among postmenopausal women should monitor for urgency UI as a possible adverse effect.
Supplementary Material
Funding
This work was supported by grants to S.R.B. from the National Institute of Diabetes, Digestive, and Kidney Disorders (grant number 1K12DK111028), the National Institute on Aging (grant number 1R03AG067937), and the UCSF Claude D. Pepper Older Americans Independence Center funded by National Institute on Aging (grant number P30 AG044281), the Helen Diller Family Chair in Population Science for Urologic Cancer at the University of California, San Francisco to S.A.K., the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services (contracts HHSN268201600018C, HHSN268201600001C, HHSN268201600002C, HHSN268201600003C, and HHSN268201600004C), and resources through the VA Puget Sound Health Care System, Seattle, Washington to M.S. The study funders had no role in the design, methods, subject recruitment, data collections, analysis, or preparation of this paper.
Conflict of Interest
None declared.
Author Contributions
S.R.B.: conception and design, acquisition of data, analysis and interpretation of data, drafting and revising the article, final approval of the version to be published. M.S.: conception and design, acquisition of data, analysis and interpretation of data, revising the article for important intellectual content, final approval of the version to be published. S.A.K., L.L.S., A.M.S., A.J.P., L.R.G., B.C., C.I., M.G.: analysis and interpretation of data, revising the article for important intellectual content, final approval of the version to be published. C.H.: acquisition of data, analysis and interpretation of data, revising the article for important intellectual content, final approval of the version to be published. B.N.B.: conception and design, acquisition of data, analysis and interpretation of data, revising the article for important intellectual content, final approval of the version to be published.
References
- 1. Gorina Y, Schappert S, Bercovitz A, Elgaddal N, Kramarow E. Prevalence of incontinence among older Americans. Vital Health Stat 3. 2014;3( 36):1–33. https://www.cdc.gov/nchs/data/series/sr_03/sr03_036.pdf [PubMed] [Google Scholar]
- 2. Wu JM, Hundley AF, Fulton RG, Myers ER. Forecasting the prevalence of pelvic floor disorders in U.S. Women: 2010 to 2050. Obstetrics and gynecology. 2009;114(6):1278–1283. doi: 10.1097/AOG.0b013e3181c2ce96 [DOI] [PubMed] [Google Scholar]
- 3. Coyne KS, Wein A, Nicholson S, Kvasz M, Chen CI, Milsom I. Economic burden of urgency urinary incontinence in the United States: a systematic review. J Manag Care Pharm. 2014;20:130–140. doi: 10.18553/jmcp.2014.20.2.130 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Huang AJ, Brown JS, Kanaya AM, et al. Quality-of-life impact and treatment of urinary incontinence in ethnically diverse older women. Arch Intern Med. 2006;166:2000–2006. doi: 10.1001/archinte.166.18.2000 [DOI] [PubMed] [Google Scholar]
- 5. Brown JS, Vittinghoff E, Wyman JF, et al. Urinary incontinence: does it increase risk for falls and fractures? Study of osteoporotic fractures research group. J Am Geriatr Soc. 2000;48:721–725. doi: 10.1111/j.1532-5415.2000.tb04744.x [DOI] [PubMed] [Google Scholar]
- 6. Tinetti ME, Inouye SK, Gill TM, Doucette JT. Shared risk factors for falls, incontinence, and functional dependence. Unifying the approach to geriatric syndromes. JAMA. 1995;273:1348–1353. doi: 10.1001/jama.1995.03520410042024 [DOI] [PubMed] [Google Scholar]
- 7. Danforth KN, Townsend MK, Lifford K, Curhan GC, Resnick NM, Grodstein F. Risk factors for urinary incontinence among middle-aged women. Am J Obstet Gynecol. 2006;194:339–345. doi: 10.1016/j.ajog.2005.07.051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Townsend MK, Curhan GC, Resnick NM, Grodstein F. BMI, waist circumference, and incident urinary incontinence in older women. Obesity (Silver Spring). 2008;16:881–886. doi: 10.1038/oby.2008.14 [DOI] [PubMed] [Google Scholar]
- 9. 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):346–353. doi: 10.1097/01.AOG.0000270121.15510.57 [DOI] [PubMed] [Google Scholar]
- 10. Subak LL, King WC, Belle SH, et al. Urinary incontinence before and after bariatric surgery. JAMA Intern Med. 2015;175:1378–1387. doi: 10.1001/jamainternmed.2015.2609 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Subak LL, Wing R, West DS, et al. Weight loss to treat urinary incontinence in overweight and obese women. N Engl J Med. 2009;360:481–490. doi: 10.1056/NEJMoa0806375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Townsend MK, Danforth KN, Rosner B, Curhan GC, Resnick NM, Grodstein F. Physical activity and incident urinary incontinence in middle-aged women. J Urol. 2008;179:1012–1016; discussion 1016–1017. doi: 10.1016/j.juro.2007.10.058 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Danforth KN, Shah AD, Townsend MK, et al. Physical activity and urinary incontinence among healthy, older women. Obstet Gynecol. 2007;109:721–727. doi: 10.1097/01.AOG.0000255973.92450.24 [DOI] [PubMed] [Google Scholar]
- 14. Seale JL. Energy expenditure measurements in relation to energy requirements. Am J Clin Nutr. 1995;62(5 suppl):1042S–1046S. doi: 10.1093/ajcn/62.5.1042S [DOI] [PubMed] [Google Scholar]
- 15. Ensrud KE, Cauley J, Lipschutz R, Cummings SR. Weight change and fractures in older women. Study of Osteoporotic Fractures Research Group. Arch Intern Med. 1997;157:857–863. doi: 10.1001/archinte.1997.00440290041004 [DOI] [PubMed] [Google Scholar]
- 16. Newman AB, Yanez D, Harris T, Duxbury A, Enright PL, Fried LP. Weight change in old age and its association with mortality. J Am Geriatr Soc. 2001;49:1309–1318. doi: 10.1046/j.1532-5415.2001.49258.x [DOI] [PubMed] [Google Scholar]
- 17. Subak LL, Richter HE, Hunskaar S. Obesity and urinary incontinence: epidemiology and clinical research update. J Urol. 2009;182(6 suppl):S2–S7. doi: 10.1016/j.juro.2009.08.071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Tennstedt SL, Link CL, Steers WD, McKinlay JB. Prevalence of and risk factors for urine leakage in a racially and ethnically diverse population of adults: the Boston Area Community Health (BACH) Survey. Am J Epidemiol. 2008;167:390–399. doi: 10.1093/aje/kwm356 [DOI] [PubMed] [Google Scholar]
- 19. Dallosso H, Matthews R, McGrother C, Donaldson M. Diet as a risk factor for the development of stress urinary incontinence: a longitudinal study in women. Eur J Clin Nutr. 2004;58:920–926. doi: 10.1038/sj.ejcn.1601913 [DOI] [PubMed] [Google Scholar]
- 20. Maserejian NN, Giovannucci EL, McVary KT, McGrother C, McKinlay JB. Dietary macronutrient and energy intake and urinary incontinence in women. Am J Epidemiol. 2010;171:1116–1125. doi: 10.1093/aje/kwq065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Bø K, Ellstrøm Engh M, Hilde G. Regular exercisers have stronger pelvic floor muscles than nonregular exercisers at midpregnancy. Am J Obstet Gynecol. 2018;218:427.e1–427.e5. doi: 10.1016/j.ajog.2017.12.220 [DOI] [PubMed] [Google Scholar]
- 22. Hendrix SL, Cochrane BB, Nygaard IE, et al. Effects of estrogen with and without progestin on urinary incontinence. JAMA. 2005;293:935–948. doi: 10.1001/jama.293.8.935 [DOI] [PubMed] [Google Scholar]
- 23. Design of the Women’s Health Initiative Clinical Trial and Observational Study. The Women’s Health Initiative Study Group. Control Clin Trials. 1998;19(1):61–109. doi: 10.1016/S0197-2456(97)00078-0 [DOI] [PubMed] [Google Scholar]
- 24. Sorensen MD, Chi T, Shara NM, et al. Activity, energy intake, obesity, and the risk of incident kidney stones in postmenopausal women: a report from the Women’s Health Initiative. J Am Soc Nephrol. 2014;25:362–369. doi: 10.1681/ASN.2013050548 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Manson JE, Greenland P, LaCroix AZ, et al. Walking compared with vigorous exercise for the prevention of cardiovascular events in women. N Engl J Med. 2002;347:716–725. doi: 10.1056/NEJMoa021067 [DOI] [PubMed] [Google Scholar]
- 26. Langer RD, White E, Lewis CE, Kotchen JM, Hendrix SL, Trevisan M. The Women’s Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol. 2003;13(9 suppl):S107–S121. doi: 10.1016/s1047-2797(03)00047-4 [DOI] [PubMed] [Google Scholar]
- 27. Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP, Agurs-Collins T. Measurement characteristics of the Women’s Health Initiative food frequency questionnaire. Ann Epidemiol. 1999;9:178–187. doi: 10.1016/s1047-2797(98)00055-6 [DOI] [PubMed] [Google Scholar]
- 28. Neuhouser ML, Tinker L, Shaw PA, et al. Use of recovery biomarkers to calibrate nutrient consumption self-reports in the Women’s Health Initiative. Am J Epidemiol. 2008;167:1247–1259. doi: 10.1093/aje/kwn026 [DOI] [PubMed] [Google Scholar]
- 29. Jackson RA, Vittinghoff E, Kanaya AM, et al. Urinary incontinence in elderly women: findings from the Health, Aging, and Body Composition Study. Obstet Gynecol. 2004;104:301–307. doi: 10.1097/01.AOG.0000133482.20685.d1 [DOI] [PubMed] [Google Scholar]
- 30. Walter W. Nutritional Epidemiology. 2nd ed. New York: Oxford University Press; 1998. [Google Scholar]
- 31. Hu FB, Stampfer MJ, Rimm E, et al. Dietary fat and coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Am J Epidemiol. 1999;149:531–540. doi: 10.1093/oxfordjournals.aje.a009849 [DOI] [PubMed] [Google Scholar]
- 32. Brown JS, Sawaya G, Thom DH, Grady D. Hysterectomy and urinary incontinence: a systematic review. Lancet. 2000;356:535–539. doi: 10.1016/S0140-6736(00)02577-0 [DOI] [PubMed] [Google Scholar]
- 33. Khullar V, Sexton CC, Thompson CL, Milsom I, Bitoun CE, Coyne KS. The relationship between BMI and urinary incontinence subgroups: results from EpiLUTS. Neurourol Urodyn. 2014;33:392–399. doi: 10.1002/nau.22428 [DOI] [PubMed] [Google Scholar]
- 34. Waetjen LE, Liao S, Johnson WO, et al. Factors associated with prevalent and incident urinary incontinence in a cohort of midlife women: a longitudinal analysis of data: study of women’s health across the nation. Am J Epidemiol. 2007;165:309–318. doi: 10.1093/aje/kwk018 [DOI] [PubMed] [Google Scholar]
- 35. Bradley CS, Erickson BA, Messersmith EE, et al. Evidence of the impact of diet, fluid intake, caffeine, alcohol and tobacco on lower urinary tract symptoms: a systematic review. J Urol. 2017;198:1010–1020. doi: 10.1016/j.juro.2017.04.097 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Eshkoor SA, Hamid TA, Shahar S, Mun CY. Factors related to urinary incontinence among the Malaysian elderly. J Nutr Health Aging. 2017;21:220–226. doi: 10.1007/s12603-016-0779-x [DOI] [PubMed] [Google Scholar]
- 37. Dallosso HM, McGrother CW, Matthews RJ, Donaldson MM. Nutrient composition of the diet and the development of overactive bladder: a longitudinal study in women. Neurourol Urodyn. 2004;23:204–210. doi: 10.1002/nau.20028 [DOI] [PubMed] [Google Scholar]
- 38. Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986;124:17–27. doi: 10.1093/oxfordjournals.aje.a114366 [DOI] [PubMed] [Google Scholar]
- 39. Carvalhais A, Da Roza T, Vilela S, Jorge RN, Bø K. Association between physical activity level and pelvic floor muscle variables in women. Int J Sports Med. 2018;39:995–1000. doi: 10.1055/a-0596-7531 [DOI] [PubMed] [Google Scholar]
- 40. Clarkson BD, Griffiths D, Resnick NM. Do brain structural abnormalities differentiate separate forms of urgency urinary incontinence? Neurourol Urodyn. 2018;37(8):2597–2605. doi: 10.1002/nau.23591 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Kuchel GA, Moscufo N, Guttmann CR, et al. Localization of brain white matter hyperintensities and urinary incontinence in community-dwelling older adults. J Gerontol A Biol Sci Med Sci. 2009;64:902–909. doi: 10.1093/gerona/glp037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Hubeaux K, Deffieux X, Ismael SS, Raibaut P, Amarenco G. Autonomic nervous system activity during bladder filling assessed by heart rate variability analysis in women with idiopathic overactive bladder syndrome or stress urinary incontinence. J Urol. 2007;178:2483–2487. doi: 10.1016/j.juro.2007.08.036 [DOI] [PubMed] [Google Scholar]
- 43. Im HW, Kim MD, Kim JC, Choi JB. Autonomous nervous system activity in women with detrusor overactivity. Korean J Urol. 2010;51:183–186. doi: 10.4111/kju.2010.51.3.183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Morres ID, Hatzigeorgiadis A, Stathi A, et al. Aerobic exercise for adult patients with major depressive disorder in mental health services: a systematic review and meta-analysis. Depress Anxiety. 2019;36:39–53. doi: 10.1002/da.22842 [DOI] [PubMed] [Google Scholar]
- 45. Coyne KS, Kvasz M, Ireland AM, Milsom I, Kopp ZS, Chapple CR. Urinary incontinence and its relationship to mental health and health-related quality of life in men and women in Sweden, the United Kingdom, and the United States. Eur Urol. 2012;61:88–95. doi: 10.1016/j.eururo.2011.07.049 [DOI] [PubMed] [Google Scholar]
- 46. Alwis US, Delanghe J, Dossche L, et al. Could evening dietary protein intake play a role in nocturnal polyuria? J Clin Med. 2020;9(8):2532. doi: 10.3390/jcm9082532 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Bankir L, Roussel R, Bouby N. Protein- and diabetes-induced glomerular hyperfiltration: role of glucagon, vasopressin, and urea. Am J Physiol Renal Physiol. 2015;309:F2–F23. doi: 10.1152/ajprenal.00614.2014 [DOI] [PubMed] [Google Scholar]
- 48. Nappo F, Esposito K, Cioffi M, et al. Postprandial endothelial activation in healthy subjects and in type 2 diabetic patients: role of fat and carbohydrate meals. J Am Coll Cardiol. 2002;39:1145–1150. doi: 10.1016/s0735-1097(02)01741-2 [DOI] [PubMed] [Google Scholar]
- 49. Tentolouris N, Tsigos C, Perea D, et al. Differential effects of high-fat and high-carbohydrate isoenergetic meals on cardiac autonomic nervous system activity in lean and obese women. Metabolism. 2003;52:1426–1432. doi: 10.1016/s0026-0495(03)00322-6 [DOI] [PubMed] [Google Scholar]
- 50. LaMonte MJ, Lee I-M, Rillamas-Sun E, et al. Comparison of questionnaire and device measures of physical activity and sedentary behavior in a multi-ethnic cohort of older women. J Meas Phys Behav. 2019;2(2):82–93. doi:10.1123/jmpb.2018-0057 [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.