Objective
To study associations between urinary incontinence (UI) symptoms, depression, and posttraumatic stress disorder in women veterans.
Study Design
This cross-sectional study enrolled women 20 to 52 years of age registered at 2 midwestern US Veterans Affairs Medical Centers or outlying clinics within 5 years preceding study interview. Participants completed a computer-assisted telephone interview assessing urogynecologic, medical, and mental health. Multivariable analyses studied independent associations between stress and urgency UI and depression and posttraumatic stress disorder.
Results
Nine hundred sixty-eight women mean aged 38.7 ± 8.7 years were included. Of these, 191 (19.7%) reported urgency/mixed UI and 183 (18.9%) stress UI. Posttraumatic stress disorder (odds ratio, 1.8; 95% confidence interval, 1.0–3.1) but not depression (odds ratio, 1.2; 95% confidence interval, 0.73–2.0) was associated with urgency/mixed UI. Stress UI was not associated with posttraumatic stress disorder or depression.
Conclusion
In women veterans, urgency/mixed UI was associated with posttraumatic stress disorder but not depression.
Psychologic distress, anxiety and depressive symptoms have been linked to urinary incontinence (UI) in both clinic-based studies and larger epidemiologic studies of middle-aged and older adults. In particular, some studies have demonstrated associations between urgency UI (but not stress UI) and depression and anxiety, leading to hypotheses that these conditions and urgency UI may share common neuropharmacologic pathways.
Mental health disorders, including depression and posttraumatic stress disorder (PTSD) (an anxiety disorder that develops in response to a traumatic experience), are particularly prevalent in veteran populations. However, UI and associated risk factors have not been well-studied in women veterans, and previous studies examining associations between anxiety disorders and UI have not focused on PTSD. The objective of this study was to measure the 12-month period prevalence of UI symptoms in reproductive-aged women veterans and to study associations between stress and urgency UI symptoms and psychologic symptoms, especially those related to depression and PTSD.
Materials and Methods
This article describes a prospectively planned secondary analysis from a cross-sectional study of reproductive-aged women veterans who registered for health care or other veteran services through the Iowa City or Des Moines VA Medical Centers or outlying clinics during the 5 years preceding study interviews (performed July 2006-Sept. 2008). The study’s primary aim was to determine whether the odds of gynecologic disorders (especially cervical cytologic abnormalities and cervical cancer) were greater for sexually assaulted women veterans in comparison to nonassaulted peers. This study was approved by the University of Iowa Institutional Review Board and the Iowa City VA Research and Development Committee. All participants signed an informed consent document.
Adult women veterans ≤52 years of age were identified using the VistA System (Veterans Health Information Systems and Technology Architecture) and recruited via mail and telephone. Women older than 52 years were not recruited and those aware of an in utero dethylstilbesterol exposure or currently receiving immunosuppressant therapy were excluded, related to the aims of the primary study. Eligible women who agreed to participate completed a computer-assisted telephone interview (CATI) administered by a trained female interviewer. Data collection included multiple validated instruments and items intended to screen for a history of sexual violence as well as urogynecologic, medical, and mental health disorders.
We identified UI symptoms using 3 standardized epidemiologic items previously used in other large epidemiologic studies. These items queried whether UI symptoms occurred during the past 12 months, including stress UI (urinary leakage “with an activity like coughing, lifting, or exercise”), urgency UI (leakage “with an urge or pressure to urinate and you couldn’t get to the toilet fast enough”) and “other” UI (leakage “without an activity like coughing, lifting, or exercise OR an urge to urinate”) symptoms. Subjects who reported 1 or more types of UI were asked to quantify the UI frequency (separately for each type) as “every day,” “a few times a week.” “a few times a month,” or “a few times a year.” Participants reporting UI symptoms were also asked how much the leakage bothered them, from “not at all” to “greatly.”
We assessed current depression and PTSD with validated diagnostic instruments based on Diagnostic and Statistical Manual of Mental Disorders–4th edition (DSM-IV) criteria. Depression was defined using the Composite International Diagnostic Interview-Short Form for Major Depression (CIDI-SF), which identifies depressive symptoms occurring during the past 12 months. The diagnosis of depression requires (1) depressed mood or anhedonia, for most of each day for most of a 2-week period, and (2) the presence of at least 4 other symptoms. CIDI-SF depression diagnoses have a sensitivity of 90% and specificity of 94% compared with the full-length CIDI.
PTSD was identified using the self-report Posttraumatic Symptom Scale (PSS-I), shown to have excellent sensitivity (88%) and specificity (96%) to structured clinical interview (SCID) diagnoses of PTSD. The assessment of lifetime traumatic events was augmented with combat-related traumatic exposures (for those who had served in combat). Follow-up questions regarding duration of symptoms and the effects of symptoms on daily functioning were also asked. Diagnosis of PTSD was determined (according to DSM-IV Criteria) if participants identified a traumatic exposure; endorsed at least 1 intrusive recollection, 3 avoidance actions, and 2 hyperarousal symptoms; acknowledged symptom duration lasting more than 1 month; and had symptoms that interfered with daily functioning. Both the CIDI-SF and PSS-I have been successfully administered in previous studies using telephone interviews.
Sexual violence exposures were assessed using detailed questions developed from the National Violence Against Women Survey (NVAWS), the National Women’s Study (NWS), and a survey of rape in Iowa. Women were asked a minimum of 5 questions that focused on whether they had experienced different types of assault throughout their lifetime. Sexual assault was defined as any act occurring without a women’s consent involving the use or threat of force, and including completed sexual penetration of the vagina, mouth, or rectum.
Generic health-related quality of life (QOL) was assessed using the Short Form, 12-Item General Health Survey (SF-12). Participants were asked to self-report height and weight, and provided information about other health behaviors and risks such as smoking, caffeine intake, exercise, and drug and alcohol use. Participants were also asked whether they had taken any medication during the past 6 months for anxiety or depression. A history of head injury was assessed by asking, “Have you ever had a head injury, skull fracture, or concussion?” Additional information collected included: demographics, military service details, past reproductive history and medical diagnoses.
Stress and urgency UI were defined as the presence of stress and urgency UI symptoms (respectively) that occurred a “few times monthly” or more often. Women who met criteria for both stress and urgency UI were classified as having mixed UI. Because the urgency UI group was small, the urgency and mixed UI groups were combined for all analyses. Risk factors, associated conditions, and other characteristics were compared between women with urgency/mixed UI and women with no UI symptoms, and similar comparisons were made in women with stress UI to the no UI group. Women with minimal UI symptoms (occurring “a few times a year” only) were not included in the comparative analyses.
Bivariable comparisons were performed using Pearson’s χ 2 or Fisher exact tests for categorical variables and the Student t test for continuous variables. Multivariable logistic regression analyses were performed, again among women with urgency/mixed UI and no UI, and among women with stress UI and no UI. The dependent variable in each model was urgency/mixed UI or stress UI. Categorization of the independent variables was examined during bivariable analysis, and best function forms were applied to the fitted multivariable models.
Several preliminary models for each dependent variable were initially created, each including related independent variables (demographic, general health/health risk, reproductive health, and mental health variables). Variables associated with the dependent outcomes in the bivariable analyses (at P < .2) were included in preliminary models. Those that remained associated with the dependent outcomes (at P < .1) in the preliminary models in addition to the primary independent variables (depression and PTSD) were included in the final models. Variables that became nonsignificant when placed in the final models were removed if no significant changes were noted in the estimates for the other variables after removing them. Odds ratios with 95% confidence intervals for each of the significant risk factors were computed from the fitted models. One-way interaction terms for sexual assault, depression, and PTSD were created and tested in the final model-building process, but these were not significant and were left out of the final models.
Statistical analyses were performed using SAS 9.1 (SAS Institute, Inc, Cary, NC). Associations were considered significant at a P < .05 level.
Results
Of 1670 women contacted, 1055 (63%) agreed to participate and 1004 completed data collection (17 were ineligible, 21 could not be reached by telephone and 13 were unable to complete the interview). Those who refused participation were similar to participants in age, self-reported general health, and number of gynecologic health care visits in the past year. The average study interview length was 1 hour and 16 minutes, and 89% were completed in a single call. After excluding 36 participants with current or recent pregnancy, 968 were included in this analysis.
The mean (± standard deviation [SD]) age was 38.7 ± 8.7 years (range, 20–52) and body mass index (BMI) 28.2 ± 6.1 kg/m 2 . Median parity was 2 (0–6). Most (60.1%) had served in active regular military duty, 11.7% were Reserve or National Guard, and 28.2% reported both service types. The average length of active duty was 5.4 ± 5.0 years. Thirty-six percent reported currently smoking, and 32% reported some type of chronic medical problem, most frequently diabetes (3.9%), and hypertension (5.4%). Past sexual assault was reported by 495 (51.1%) of the women veterans, with 329 (34.0%) reporting more than 1 past assault.
Depression and PTSD were common, occurring in 301 (31%) and 242 (25%) women, respectively. Fifteen percent met criteria for both depression and PTSD. Overall, 448 (46.3%) reported taking a medication for anxiety or depression during the past 6 months, including 81% of those with depression and 73% of those with PTSD.
UI symptoms were common ( Table 1 ) and bothersome. Stress and/or urgency UI occurred at least “a few times a month” in 374 women (38.6%) during the past year. One hundred thirty-two women (13.6%), 99 (10.2%), and 108 (11.2%) reported they were “somewhat,” “very much,” and “greatly” bothered, respectively, by their urinary leakage.
| Variable | n (%) | “Few times a month” | “Few times a week” | “Every day” |
|---|---|---|---|---|
| Stress UI a | 183 (18.9) | 101 (10.4) | 59 (6.1) | 23 (2.4) |
| Mixed UI b | 157 (16.2) | 25 (2.6) | 57 (5.9) | 75 (7.7) |
| Urgency UI c | 34 (3.5) | 22 (2.3) | 8 (0.8) | 4 (0.4) |
| Other UI d | 13 (1.3) | |||
| Minimal UI e | 245 (25.3) | |||
| No UI | 334 (34.5) | |||
| Missing data | 2 (0.2) | |||
| Total | 968 (100.0) |
a UI with activity, occurring at least “a few times a month;”
b UI with activity and with urge to urinate, both occurring at least “a few times a month;”
c UI with urge to urinate, occurring at least “a few times a month;”
d UI without activity or urge to urinate, occurring at least “a few times a month;”
Characteristics of women with no UI, stress UI, and urgency/mixed UI are presented in Table 2 . In bivariable analyses, compared with women with no UI, women with stress UI and women with urgency/mixed UI were older, more often married, less likely to be a student, and more likely to be unemployed. Women with stress UI and those with urgency/mixed UI (compared with women with no UI) were also heavier, more likely to report chronic medical problems, more parous, more often menopausal, and were more likely to have had hysterectomy and a history of urinary tract infection.
| Population characteristics | No UI (n = 334) | Stress UI (n = 183) | P value a | Urgency/mixed UI (n = 191) | P value a |
|---|---|---|---|---|---|
| Demographics | |||||
| Age, y | < .0001 | < .0001 | |||
| ≥45 | 74 (22.2) | 65 (35.5) | 96 (50.3) | ||
| 40-44 | 48 (14.4) | 43 (23.5) | 42 (22.0) | ||
| 30-39 | 95 (28.4) | 53 (29.0) | 39 (20.4) | ||
| <30 | 117 (35.0) | 22 (12.0) | 14 (7.3) | ||
| Race | .11 | .06 | |||
| Nonwhite | 42 (12.6) | 15 (8.2) | 12 (6.3) | ||
| White | 255 (76.4) | 154 (84.2) | 153 (80.1) | ||
| Multirace | 37 (11.1) | 14 (7.7) | 26 (13.6) | ||
| Education | < .99 | .52 | |||
| High School/GED completion | 53 (15.9) | 29 (15.9) | 33 (17.3) | ||
| Some college/technical training | 180 (53.9) | 99 (54.1) | 112 (58.6) | ||
| College completed or greater | 101 (30.2) | 55 (30.1) | 46 (24.1) | ||
| Employment | .03 | < .0001 | |||
| Employed | 161 (48.2) | 88 (48.1) | 94 (49.2) | ||
| Retired | 4 (1.2) | 3 (1.7) | 17 (8.9) | ||
| Student | 108 (32.3) | 41 (22.4) | 25 (13.1) | ||
| Unemployed | 61 (18.3) | 51 (27.9) | 55 (28.8) | ||
| Marital status | .0001 | < .0001 | |||
| Single | 101 (30.2) | 31 (17.0) | 24 (12.6) | ||
| Divorced | 104 (31.1) | 56 (30.6) | 75 (39.3) | ||
| Married | 129 (38.7) | 96 (52.5) | 92 (48.2) | ||
| Combat or war zone exposure | 108 (32.3) | 55 (30.1) | .59 | 40 (20.9) | .005 |
| General health/risks | |||||
| Body mass index, kg/m 2 | 26.6 ± 5.3 | 28.8 ± 5.9 | .0001 | 30.6 ± 7.2 | < .0001 |
| Exercise, min per wk | .10 | .0002 | |||
| 0-59 | 118 (35.4) | 81 (45.0) | 95 (49.7) | ||
| 60-120 | 88 (26.4) | 38 (21.1) | 55 (28.8) | ||
| >120 | 127 (38.1) | 61 (33.9) | 41 (21.5) | ||
| Caffeine (no. of beverages/d) | 2.4 ± 2.8 | 2.6 ± 2.8 | .47 | 3.2 ± 3.9 | .03 |
| Chronic medical problem(s) (yes/no) | 76 (22.8) | 70 (38.3) | .0002 | 75 (39.3) | < .0001 |
| Head injury | 97 (29.0) | 59 (32.2) | .45 | 86 (45.0) | .0002 |
| Currently smoke | 106 (31.7) | 58 (31.7) | .99 | 80 (41.9) | .02 |
| Reproductive health | |||||
| Parity | < .0001 | < .0001 | |||
| 0 | 163 (48.8) | 44 (24.0) | 47 (24.6) | ||
| 1 | 65 (19.5) | 40 (21.9) | 38 (19.9) | ||
| 2 | 73 (21.9) | 57 (31.2) | 56 (29.3) | ||
| ≥3 | 33 (9.9) | 42 (23.0) | 50 (26.2) | ||
| Menopause | 66 (20.1) | 49 (28.3) | .04 | 77 (41.9) | < .0001 |
| Hysterectomy | 36 (10.8) | 35 (19.1) | .008 | 59 (30.9) | < .0001 |
| History of UTI | 143 (42.9) | 100 (54.6) | .01 | 142 (74.4) | < .0001 |
| Mental health/risks | |||||
| History of sexual assault | 134 (40.1) | 98 (53.6) | .003 | 123 (64.4) | < .0001 |
| Number of sexual assaults | .0003 | < .0001 | |||
| None | 200 (59.8) | 85 (46.5) | 68 (35.6) | ||
| Single assault | 61 (18.3) | 28 (15.3) | 27 (14.1) | ||
| >1 assault | 73 (21.9) | 70 (38.3) | 96 (50.3) | ||
| Medication use for depression or anxiety in last 6 mo | 115 (34.4) | 94 (51.4) | .0002 | 119 (62.3) | < .0001 |
| Depression | 79 (23.7) | 60 (32.8) | .03 | 83 (43.5) | < .0001 |
| PTSD | 68 (20.4) | 45 (24.6) | .27 | 71 (37.2) | < .0001 |
a P values indicate comparison with “No UI” group using Pearson’s χ 2 , Fisher exact, or Student t test.
Other variables differed in their associations with urgency/mixed UI and with stress UI. In bivariable analyses, women with urgency/mixed UI more often smoked, exercised less, and more often reported a history of head injury. A history of sexual assault and depression were each more common in women with either stress UI or urgency/mixed UI. In contrast, PTSD was identified more often in women with urgency/mixed UI compared with women with no UI (37% vs 20%), but not in women with stress UI only (25%). Both women with urgency/mixed UI and those with stress UI were more likely to report taking a medication for anxiety or depression in the past 6 months than were women with no UI.
The higher rates of mental health disorders in the urgency/mixed UI group were reflected in lower mental health related QOL (SF-12 mental component scores 44.6 ± 9.0 vs 47.2 ± 7.8 for urgency/mixed vs no UI, P ≤ .001). SF-12 mental component scores were similar in the stress UI and no UI groups ( P = .57). SF-12 physical component scores were also lower in women with urgency/mixed UI (39.9 ± 6.9 vs 42.0 ± 6.7, P = .001) and stress UI (40.7 ± 7.3 vs 42.0 ± 6.7, P = .04), both compared with women with no UI.
Tables 3 and 4 present the multivariable analyses results. Depression and PTSD were not independently associated with stress UI ( Table 3 ). Variables independently associated with stress UI included age, BMI, race, parity, a history of chronic medical problems, and a history of sexual assault. In contrast, PTSD was independently associated with urgency/mixed UI, but depression was not ( Table 4 ). Other variables independently associated with urgency/mixed UI included age, BMI, race, exercise, parity, a history of UTI, and a history of sexual assault.
