Volume 201, Issue 5 , Pages 490.e1-490.e7, November 2009
Major depression and urinary incontinence in women: temporal associations in an epidemiologic sample
Article Outline
Objective
To determine whether: (1) major depression is associated with increased risk for onset of urinary incontinence, and (2) urinary incontinence is associated with increased risk for onset of depression.
Study Design
Longitudinal cohort study of female Health and Retirement Study participants completing baseline interviews at Wave 3 (1996-1997) and follow-up interviews at Waves 4-6 (1998-2003).
Results
In a cohort of 5820 women with a mean age 59.3 (± 0.5) years, 6-year cumulative incidences of depression and incontinence were 11% and 21%, respectively. Major depression was associated with increased odds of incident incontinence (adjusted odds ratio, 1.46; 95% confidence interval, 1.08–1.97) during follow-up compared with those without major depression at baseline. Conversely, incontinence was not associated with increased odds of incident depression (adjusted odds ratio, 1.03; 95% confidence interval, 0.75–1.42) compared with those without incontinence at baseline.
Conclusion
Major depression predicted onset of urinary incontinence in a population-based sample of at-risk, community-dwelling women. Incontinence did not predict onset of depression.
Key words: incidence, longitudinal, major depression, urinary incontinence
Major depression and urinary incontinence (UI) are prevalent, distressing disorders that disproportionately affect women, with approximately one-fifth who have major depression1 and nearly one-half who have incontinence2 during their lifetimes. These illnesses are of major public health importance because of the high number of individuals affected, adverse effects on functioning, and quality of life (QOL) decrements.3, 4 This public health impact will only intensify with the aging of the US population, as rates of UI increase dramatically with age.5 There is now ample cross-sectional evidence that UI and depression often co-occur in women. Women with both disorders report significantly greater decrements in QOL and functional status and increased incontinence symptom burden compared with women with UI alone.6, 7
The association between these disorders has been examined in clinical and population-based cross-sectional samples. Depression has been assessed using measures of depressive symptoms8, 9, 10, 11, 12, 13 as well as criteria for major depressive disorder.5, 6, 14, 15, 16 Initially, reports revealed high rates of major depression or depressive symptoms in women with UI attending specialty clinics, with the highest rates in women with urge incontinence.6, 9, 14 This association was confirmed in 2 large US population-based studies showing that women with major depression had 1.8 (range, 1.3–2.6) to 2.5 (range, 1.7–3.7) odds of UI, compared with nondepressed women.5, 15 A third large Canadian population-based study found similarly increased odds (odds ratio [OR], 5.7; 95% confidence interval [CI], 3.1–10.5), although the very low rate of UI in this sample (3.2%) complicates comparison with the other published studies.16
This cross-sectional relationship between depression and UI could be explained by common neurologic or biochemical pathways underlying both conditions. The increased activity of the hypothalamic-pituitary axis seen in depression could also result in physiologic changes contributing to involuntary urine loss. Alternatively, the chronic embarrassment, social isolation, and symptom burden associated with UI could lead to depression over time.
These temporal or causal relationships remain unclear. Well-controlled longitudinal studies examining incidence for onset of comorbid disease in at-risk populations are needed. To address these questions, we examined a longitudinal population-based cohort to determine whether: (1) major depression is associated with increased risk for onset of UI, and (2) UI is associated with increased risk for onset of depression.
Materials and Methods
Sample
The study sample included participants from the Health and Retirement Study (HRS), a prospective population-based cohort study funded by the National Institute on Aging (NIA) and conducted by the University of Michigan. The HRS provides longitudinal data on health, financial, and sociologic issues surrounding retirement.17 At the initial interview in 1991, participants ranged in age from 51-61 years; spouses of any age were also interviewed. To identify age-eligible persons, the HRS screened approximately 70,000 households obtained from an area probability sample.18 The baseline response rate was 82%, and comparison with the 1990 US Census data indicated no sample bias.19 Participants complete biennial telephone interview “waves.” The HRS is under current Institutional Review Board approval by the University of Michigan and NIA.
This study used data from Waves 3 (1996-97) through Wave 6 (2002-03), because specific UI questions were first included at Wave 3. The sample was drawn from all female respondents at Wave 3. Exclusion criteria were institutionalized participants, interviews obtained via proxy, and those with missing observations at Wave 3. A previous analysis revealed a baseline cross-sectional association between UI and depression for female respondents at Wave 3.15
Measures and data collection
UI was assessed at each wave. Participants endorsing the question, “During the last 12 months, have you lost any amount of urine beyond your control?” were then asked, “On about how many days in the last month have you lost any urine?” Women reporting urine loss in the past year and ≥1 day of incontinence in the past month were classified as incontinent. The International Continence Society criteria for diagnosing UI rely on symptoms, no longer requiring an examination, and presence of involuntary urine loss in the past month, rather than the past year, has become the standard in large epidemiologic studies as a more valid, reliable, and sensitive indicator of the functional impact of incontinence.5, 20, 21 Although frequency of urine loss was assessed, amount was not, precluding assessment of UI severity (ie, calculation of a severity index).
Measures of depressive symptomatology were included at all waves. A short form of the Composite International Diagnostic Interview (CIDI-SF)22 was introduced at Wave 3 and administered only once to each participant during her HRS involvement. Major depression on the CIDI-SF is diagnostic, using the gold standard, DSM-IV-TR diagnostic criteria, which requires endorsement of ≥5 symptoms, including depressed mood or anhedonia, for most days of a 2-week duration in the past year. In addition, an 8-item version of the Center for Epidemiologic Studies Depression Scale (CES-D)23 was administered at each wave to assess depressive symptoms during the week preceding the interview. Endorsing ≥6 of the 8 items on this CES-D version is compatible with a high likelihood of a diagnosis of major depression. This cut-point is based on the established cut-point of ≥16 on the full scale24 and has been used previously in this population.15, 25 A comparison of individuals scoring positive on the 8-item CES-D and CIDI-SF in this population showed roughly equivalent rates of self-report of psychiatric diagnosis or psychiatric treatment, with the CES-D having a stronger association with antidepressant use.25
All waves collected the following self-reported covariates: age, race, education, marital status, employment, household income, medical conditions, functional status, body mass index (BMI), exercise, smoking, alcohol, psychiatric medications, and parity. Specific medical conditions included hypertension, diabetes, cancer, lung disease, heart conditions, stroke, and arthritis. Number of medical conditions was used to create a comorbidity index with categories: none (0), mild (1-2), and moderate/severe (3+). Functional status was determined using a group of 10 functional limitations questions and 6 activities of daily living questions, with responses categorized into 4 levels based on degree of limitations: no impairments, greater mobility/impairment, moderate mobility/strength impairment, and activities of daily living impairment. These categories rank items based on the proportion of older HRS participants' reports of difficulty for each question.26, 27 BMI, defined as kg/m2, was categorized: normal (<25), overweight (25-29.9), and obese (≥30). Dichotomous variables were as follows: regular exercise (vigorous physical activity ≥3 times per week vs other), smoking (current vs other), alcohol use (current vs other), and psychiatric medication use (current vs other). Parity was entered as a single categorical variable (0, 1-2, and ≥3 deliveries). Additional information regarding obstetric history was not obtained.
Statistical analyses
Two analyses were conducted to examine the relationship between major depression and UI. In the first, depression at Wave 3 (baseline) was used to predict incident UI in follow-up waves (4-6). Participants with UI (ie, ≥1 day in the past month) at Wave 3 were considered not at risk for incident UI developing and were excluded from this analysis. Women with major depression on the CIDI-SF at Wave 3 were considered cases; those without major depression were considered controls. Participants' responses to questions about UI at follow-up waves were used to determine whether UI had developed.
In the second analysis, UI at Wave 3 (baseline) was used to predict incident depression in follow-up waves (4-6). Participants with probable major depression on the CES-D (ie, ≥6 of 8 items positive) at Wave 3 were considered not at risk for incident depression developing and were excluded from this analysis. Women with UI at Wave 3 were considered cases; those without UI were considered controls. Participants' responses to the CES-D at follow-up waves were used to determine whether probable major depression had developed. The CES-D was used as the depressive measure for this incidence analysis as it was the only measure administered to all participants at all waves.
The number of new cases of UI and depression among individuals identified as at risk (at baseline) are reported as 6-year cumulative incidences. Reporting the condition at any time during follow-up was counted as an incident condition. Univariate analyses (using χ2 tests, independent t test, or analysis of variance) were conducted to investigate each covariate's association with either incident UI or incident depression. Both theoretically relevant variables (eg, medical comorbidity) and variables associated with incident UI or incident depression on univariate analyses (P < .05) were included in multivariable logistic regression models. We added each variable to the model with only the main predictor 1 × 1 and chose the model with the best model fitting evaluated by the Akaike's Information Criterion (AIC). If multiple variables resulted in similar AIC, we selected the variable with the most important clinical relevance. Variables not chosen in the first step were added to the model with the main predictor and the selected covariate 1 × 1, and the best model determined by AIC and/or clinical relevance was selected. The same step was repeated until the best parsimonious model was found. Because age and education were important potential confounders, they were added to the final models if they were not selected in the process. Odds ratios were derived for the probability of either incident UI or incident depression developing during follow-up, adjusting for covariates.
All statistical analyses were conducted using SAS 9.1 (SAS Institute, Cary, NC).28 We analyzed the data using SAS survey procedures (PROC Surveymeans, PROC Surveyfreq, PROC Surveyreg, and PROC Surveylogistic), which estimate sampling errors for complex sample data using a Taylor Series linearization of the estimator. All analyses incorporated sampling weights and strata provided by the HRS to account for the HRS sampling framework.
Results
The baseline sample is the 5820 eligible female participants completing Wave 3 interviews. Mean age was 59.3 years (± 0.05), with a range of 25-76 years. There was minimal attrition, with 5328 (91.5%), 5001 (85.9%), and 4768 (81.9%) participants completing interviews for Waves 4, 5, and 6, respectively. Seven-hundred eighty-four (13.5%) women reported UI at baseline and were therefore considered not at risk for having incident UI develop during follow-up. Similarly, 416 (7.2%) women had probable major depression on the CES-D at baseline and were therefore considered not at risk for incident depression developing during follow-up. Depression at baseline was not associated with loss to follow-up, including death (χ2 = 1.63; P = .202).
The 6-year cumulative incidence of UI was 21%, with 1071 new cases among 5036 women at risk. The 6-year cumulative incidence of major depression was 11%, with 590 new cases among 5404 women at risk. In the weighted analyses, 32% (27-38%) of women with depression at baseline had incident UI develop, compared with 21% (20-23%) of women without depression at baseline (χ2 = 16.787; P < .001), whereas 14% (11-17%) of women with UI at baseline had depression develop, compared with 10% (9-11%) of women without UI at baseline (χ2 = 7.992; P = .005). Table 1 shows the unweighted baseline demographic and clinical variables, as well as the baseline variables and univariate associations for the incidence analyses, incorporating the HRS sampling weights.
TABLE 1. Characteristics of respondents and univariate analyses
| Characteristic | Total (n = 5820)a | Totalb | Incident UIb | No incident UIb | P value | Incident Depb | No incident Depb | P value |
|---|---|---|---|---|---|---|---|---|
| Age (± SE) (y) | 57.5 ±0.07 | 59.3±0.05 | 59.5±0.12 | 59.3±0.07 | .091 | 58.9±0.19 | 59.4±0.06 | .020 |
| Baseline UI | 13%(n=784) | 14% | — | — | — | 18% | 13% | .005 |
| Baseline depression (CIDI-SF) | 10%(n=557) | 9% | 11% | 7% | <.001 | — | — | — |
| Nonwhite race | 21%(n=1210) | 14% | 11% | 16% | <.001 | 18% | 13% | <.001 |
| Education | ||||||||
| < High school degree | 25%(n=1450) | 24% | 23% | 24% | .865 | 38% | 20% | <.001 |
| High school degree/GED | 40%(n=2345) | 41% | 40% | 41% | 37% | 42% | ||
| Some college | 20%(n=1169) | 20% | 21% | 20% | 16% | 21% | ||
| ≥4 y college degree | 15%(n=855) | 15% | 16% | 15% | 9% | 17% | ||
| Married/partnered | 72%(n=4195) | 68% | 78% | 68% | .224 | 63% | 70% | .003 |
| Employment | <.001 | |||||||
| Employed | 50%(n=2912) | 46% | 42% | 49% | <.001 | 40% | 49% | |
| Unemployed | 14%(n=805) | 14% | 17% | 11% | 24% | 10% | ||
| Retired | 16%(n=938) | 19% | 20% | 19% | 14% | 20% | ||
| Homemaker | 20%(n=1165) | 21% | 21% | 21% | 22% | 21% | ||
| Household income | ||||||||
| <$22,000 | 32%(n=1865) | 33% | 35% | 32% | .218 | 45% | 30% | <.001 |
| $22,000-$50,999 | 34%(n=1965) | 35% | 35% | 35% | 33% | 36% | ||
| >$51,000 | 34%(n=1943) | 32% | 29% | 32% | 22% | 34% | ||
| BMI | <.001 | |||||||
| Normal (<25) | 40%(n=2264) | 41% | 35% | 45% | <.001 | 33% | 43% | |
| Overweight (25-29) | 34%(n=1918) | 34% | 38% | 33% | 33% | 34% | ||
| Obese (≥30) | 27%(n=1511) | 24% | 27% | 21% | 34% | 23% | ||
| Regular exercise (≥3 times per wk) | 47%(n=2713) | 47% | 44% | 49% | .005 | 37% | 50% | <.001 |
| Current smoker | 21%(n=1203) | 20% | 20% | 20% | .891 | 29% | 18% | <.001 |
| Current alcohol use | 46%(n=2652) | 48% | 48% | 48% | .821 | 36% | 50% | <.001 |
| Current psychiatric medication use | 8%(n=491) | 8% | 9% | 6% | <.001 | 18% | 5% | <.001 |
| Parity | .542 | |||||||
| 0 | 9%(n=516) | 8% | 7% | 9% | .010 | 8% | 9% | |
| 1-2 | 36%(n=2079) | 34% | 30% | 35% | 36% | 34% | ||
| ≥3 | 55%(n=3224) | 58% | 62% | 56% | 56% | 58% | ||
| Medical comorbidity | ||||||||
| None (0) | 26%(n=1504) | 24% | 17% | 27% | <.001 | 12% | 26% | <.001 |
| Mild (1-2) | 58%(n=3348) | 59% | 62% | 60% | 62% | 59% | ||
| Moderate-severe (≥3) | 17%(n=968) | 17% | 21% | 13% | 26% | 14% | ||
| Functional limitationsc | ||||||||
| None | 32%(n=1839) | 30% | 21% | 36% | <.001 | 18% | 34% | <.001 |
| Mild | 26%(n=1541) | 27% | 28% | 28% | — | 18% | 29% | — |
| Moderate | 30%(n=1723) | 31% | 34% | 28% | — | 43% | 29% | — |
| Severe | 12%(n=717) | 12% | 16% | 8% | — | 21% | 8% | — |
aUnweighted data; |
bincorporate HRS sampling weights; |
cfunctional limitations categories: none (no impairments), mild (greater mobility/impairment), moderate (moderate mobility/strength impairment), and severe (activities of daily living impairment). |
In multivariable logistic regression analyses, major depression (CIDI-SF) was associated with increased odds of incident incontinence (OR, 1.46; 95% CI, 1.08–1.97) during follow-up compared with those without depression at baseline, adjusting for covariates (Table 2). Other significant covariates increasing the odds of incident UI included BMI, parity (≥3 deliveries), medical comorbidity, functional limitations, and increased education, whereas nonwhite race significantly decreased the odds of incident UI. Conversely, incontinence was not associated with increased odds of incident probable major depression (CES-D) (OR, 1.03; 95% CI, 0.75–1.42) during follow-up compared with those without incontinence at baseline, adjusting for covariates (Table 3). Significant covariates increasing the odds of incident major depression included BMI, smoking, medical comorbidity, functional limitations, and psychiatric medication use, whereas age and increased education significantly decreased the odds of incident major depression. We ran a sensitivity analysis removing functional limitations from the final models, because of the potential colinearity with depression. In analysis 1, we found that our main effect, major depression, became a stronger predictor (OR, 1.63; 95% CI, 1.21–2.20) of incident UI when functional limitations were removed from the model. In analysis 2, we found that our main effect, UI, remained a nonsignificant predictor (OR, 1.14; 95% CI, 0.84–1.56) of incident depression when functional limitations were removed from the model.
TABLE 2. Adjusted odds of incident UIa
| Predictors (baseline) | Adjusted ORb (95% CI) |
|---|---|
| Major depression (CIDI-SF) | 1.46(1.08–1.97) |
| BMIc | |
| 25-29 (overweight) | 1.41(1.15–1.73) |
| ≥30 (obese) | 1.43(1.14–1.80) |
| Parityd | |
| 1-2 deliveries | 1.11(0.78–1.58) |
| ≥3 deliveries | 1.43(1.02–2.00) |
| Medical comorbiditye | |
| Mild | 1.32(1.04–1.67) |
| Moderate-severe | 1.58(1.16–2.15) |
| Functional limitationsf | |
| Mild | 1.52(1.19–1.94) |
| Moderate | 1.78(1.39–2.29) |
| Severe | 3.02(2.19–4.17) |
| Nonwhite race | 0.57(0.45–0.71) |
| Age, y | 1.01(0.98–1.04) |
| Educationg | |
| High school degree/GED | 1.19(0.95–1.49) |
| Some college | 1.45(1.11–1.88) |
| ≥4 y college degree | 1.65(1.23–2.23) |
| Psychiatric medication use | 1.05(0.75–0.46) |
aIncorporates HRS sampling weights; |
badjusting for listed covariates. Reference groups: |
cBMI <25; |
dnulliparity (no deliveries); |
eno medical comorbidity; |
fno functional limitations; |
g< high school degree. |
TABLE 3. Adjusted odds of probable major depression (CES-D ≥6)a
| Predictors (baseline) | Adjusted ORb (95% CI) |
|---|---|
| Incontinence | 1.03(0.75–1.42) |
| BMIc | |
| 25-29 (overweight) | 1.15(0.87–1.53) |
| ≥30 (obese) | 1.46(1.08–1.98) |
| Medical comorbidityd | |
| Mild | 1.65(1.15–2.36) |
| Moderate-severe | 1.74(1.11–2.74) |
| Functional limitationse | |
| Mild | 0.90(0.63–1.29) |
| Moderate | 1.85(1.34–2.57) |
| Severe | 2.17(1.42–3.32) |
| Current smoker | 1.51(1.15–1.98) |
| Age, y | 0.96(0.92–0.99) |
| Employmentf | |
| Unemployed | 1.35(0.95–1.92) |
| Retired | 0.80(0.55–1.15) |
| Homemaker | 1.03(0.75–1.40) |
| Educationg | |
| High school degree/GED | 0.63(0.48–0.83) |
| Some college | 0.58(0.41–0.82) |
| ≥4 y college degree | 0.55(0.36–0.85) |
| Psychiatric medication use | 2.63(1.84–3.77) |
aIncorporates HRS sampling weights; |
badjusting for listed covariates. Reference groups: |
cBMI <25; |
dno medical comorbidity; |
eno functional limitations; |
femployed; |
g< high school degree. |
Comment
Major depression and UI are prevalent, distressing conditions in women, and it is known that comorbid disease is common and leads to increased symptom burden and distress. However, temporal relationships between the disorders are not understood. Our longitudinal study's results suggest that major depression predicts incident UI in at-risk women. Women with major depression at baseline were 50% more likely to have UI develop during follow-up than women who were not depressed. Evidence of the opposite relationship is not supported by our study.
Identifying women at risk for comorbid disease is paramount. In the recent, large WHO World Health Survey (n = 245,404), individuals with depression had greater decrements in health than those with chronic diseases such as angina and diabetes.29 Moreover, those with comorbid depression and ≥1 chronic disease had the worst health scores of all disease states. This emphasizes the importance of addressing depression urgently as a public-health priority, not only to reduce disability from depression, but also to reduce additive disability when depression is comorbid with other chronic diseases. Women with comorbid depression and UI have markedly increased symptom burden from their disease and may be less likely to pursue diagnosis and treatment because of anhedonia and decreased self-efficacy. It is currently unknown whether successfully treating depression will decrease the risk of incident UI, but it is hopeful that it may.
Strengths of our study include using a large, population-based sample with minimal attrition over the 6-year study period. Frequent monitoring for development of incident conditions using interviews rather than medical diagnoses strengthens our results, because many women with UI and depression delay or never seek care, or have missed diagnoses. In addition, we accurately assessed numerous confounders (eg, obesity—an established risk factor for both depression and UI) and accounted for these in our analyses. Finally, both UI and depression are dynamic conditions, in terms of natural history and treatment. This makes our analysis conservative, as individuals may change case-control groups during follow-up, leading to a decreased estimate of odds. Because current depression was our predictor, women with a history of successfully treated major depression were included in our controls, also making this a conservative analysis.
A primary limitation of the study was our population of mainly older women, which does not reveal how depression and UI may interact in younger women. In addition, we only viewed the natural history of these conditions over a 6-year period, not an entire lifetime. Also, as an incidence study, it examined only those who had the outcome develop during follow-up, excluding those with depression or UI at baseline. For medications, the information on psychiatric medications does not distinguish specific classes of drugs, and information on hormone therapy was incompletely collected. Given the mean age of 59 years at baseline (Wave 3); however, it is unlikely that many women were in the menopausal transition during the study period. Further, the survey questions do not allow distinction between UI types or severity, information on comorbid psychiatric disorders was not available, and the gold standard for depression diagnoses (CIDI interview) was not available at all waves. Although it would be preferable to have CIDI diagnoses at each wave, in this population, the CES-D and CIDI-SF rates of self-reported psychiatric diagnosis or treatment did not differ, and the CES-D actually had a stronger association with current antidepressant use.25
In the only similar study to our knowledge, Thom et al30 reported on 2 cohorts of elderly women (age ≥65 years) evaluated for incident medical record diagnoses. Women with chart diagnoses of depression at baseline had a relative risk (RR) of 1.6 (1.2-2.0) of being diagnosed with UI over a 9-year follow-up.30 They also failed to find the opposite relationship, namely, women with a chart diagnosis of UI were not at increased risk of diagnosis with depression during follow-up (RR, 0.9 [0.6-1.6]). This study was limited to individuals with medical record diagnoses—an important limitation because only half of the women with significant UI seek medical care31 and approximately half of the patients with depression in primary care are not diagnosed.32 In addition, the analyses included no potential covariates for the development of incident conditions, except age and cohort.
Our results suggest the direction of causality is from depression to UI, not vice versa. There are several mechanisms by which depression could lead to UI. First, alterations in key neurotransmitters in the brain, including decreased serotonin activity, can lead to depression.33 Serotonin also affects bladder function, as a key neurotransmitter involved in both inhibiting micturition reflex pathways34 and facilitating urethral sphincter closure.35 It could follow, therefore, that abnormalities in serotonergic pathways are linked to both depression and the regulation of voiding function.36, 37 Second, the hypothalamic-pituitary axis and the sympathetic nervous system exhibit increased activity in many depressed individuals,38 resulting in increased release of cortisol and catecholamines. The chronic, increased activity of the hypothalamic-pituitary axis seen in depression may lead to physiologic changes in the bladder contributing to involuntary urine loss. Finally, depression often starts early in life and is related to many established UI risk factors like obesity and medical comorbidity. Depression could be the common underlying mechanism for these related disorders. In fact, depression in adolescence has been associated with development of obesity in women,39 potentially representing an important mediator in the relationship between depression and UI. Controlling for obesity in our analyses may therefore have led to a conservative estimate of the effect of depression on the development of UI. In contrast, because UI did not predict incident depression, our data suggest that depression as a psychologic reaction to the adverse effects of UI on functioning is a less-likely explanation for the cross-sectional association of depression and UI. UI is associated with known decrements in disease-specific QOL such as shame and avoidance caused by UI, but this is not the same as depression and should not be confused.
With the aging of our population and the emergence of UI and depression as important illnesses in women, it is necessary to understand the relationship between these comorbid disorders. Our longitudinal study suggests that major depression predicts incident UI in at-risk, community-dwelling women. This finding can be used immediately to counsel women with depression about a potentially increased risk for the development of UI and what to do if incontinence symptoms begin to emerge. Future research should investigate the treatment of depression and its effect on incontinent episodes in women with comorbid disease.
References
- Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: results from the National Comorbidity Survey. Arch Gen Psychiatry. 1994;51:8–19
- Urinary incontinence in adults: acute and chronic management (Clinical practice guideline 2). Rockville, MD: US Dept of Health and Human Services Agency for Health Care Policy and Research; 1996;
- . Quality of life of women with urinary incontinence: further development of the incontinence quality of life instrument (I-QOL). Urology. 1999;53:71–76
- . Global burden of depressive disorders in the year 2000. Br J Psychiatry. 2004;184:386–392
- . Urinary incontinence in US women: a population-based study. Arch Intern Med. 2005;165:537–542
- . Prevalence of comorbid psychiatric illness and its impact on symptom perception, quality of life, and functional status in women with urinary incontinence. Am J Obstet Gynecol. 2002;187:80–87
- . Incontinence severity and major depression in incontinent women. Obstet Gynecol. 2005;106:585–592
- . Major depression in the elderly: a population study in Helsinki. Int Psychogeriatr. 1996;8:437–443
- . Urinary incontinence and depression. J Urol. 1999;162:82–84
- The association of depressive symptoms and urinary incontinence among older adults. J Am Geriatr Soc. 2000;48:413–416
- . The contribution of hysterectomy to the occurrence of urge and stress urinary incontinence symptoms. BJOG. 2002;109:149–154
- . Depressive symptoms in women with urinary incontinence: a prospective study. Urol Nurs. 2001;21:397–399
- . The association between chronic diseases and depressive symptomatology in older Mexican Americans. J Gerontol A Biol Sci Med Sci. 1998;53:M188–M194
- . Severe depression determines quality of life in urinary incontinent women. Neurourol Urodyn. 2003;22:563–568
- . Urinary incontinence and depression in middle-aged United States women. Obstet Gynecol. 2003;101:149–156
- . Major depression in female urinary incontinence. Psychosomatics. 2006;47:147–151
- . Institutional review board information: health and retirement study. Ann Arbor: University of Michigan; 1990;
- . Predictors of onset of and recovery from mobility difficulty among adults aged 51-61 years. Am J Epidemiol. 1998;148:63–71
- . The health and retirement study. ICPSR Bull. 1993;14:1–4
- . Is urinary incontinence a barrier to exercise in women?. Obstet Gynecol. 2005;106:307–314
- . The prevalence of urinary incontinence in women in four European countries. BJU Int. 2004;93:324–330
- . The World Health Organization Composite International Diagnostic Interview Short-Form (CIDI-SF). Intern J Meth Psychiatr Res. 1998;7:171–185
- . Use of the Center for Epidemiologic Studies-Depression Scale with older adults. Clin Gerontol. 1986;5:119–136
- . The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385–401
- . A revised CES-D measure of depressive symptoms and a DSM-based measure of major depressive episodes in the elderly. Int Psychogeriatr. 1999;11:139–148
- . Prevalence and correlates of depressive symptoms in a community sample of people suffering from heart failure. J Am Geriatr Soc. 2002;50:2003–2008
- . Combining activities of daily living with instrumental activities of daily living to measure functional disability. J Gerontol B Psychol Sci Soc Sci. 1998;53:S46–S57
- SAS [computer program]. Version 9.1. Cary, NC: SAS Institute, Inc.
- . Depression, chronic diseases, and decrements in health: results from the World Health Surveys. Lancet. 2007;370:851–858
- . Medically recognized urinary incontinence and risks of hospitalization, nursing home admission and mortality. Age Ageing. 1997;26:367–374
- . Health care discussions and treatment for urinary incontinence in U.S. women. Am J Obstet Gynecol. 2006;194:729–737
- . Evidence review: efficacy and effectiveness of antidepressant treatment in primary care. Gen Hosp Psychiatry. 2002;24:213–224
- . The neurobiology of depression. Sci Am. 1998;278:42–49
- . Depression and incontinence. World J Urol. 2001;19:351–357
- . Serotonin and noradrenaline involvement in urinary incontinence, depression and pain: scientific basis for overlapping clinical efficacy from a single drug, duloxetine. Int J Clin Pract. 2007;61:1349–1355
- . Glutamate activation of neuron in CV-reactive areas of cat brain stem affects urinary bladder mobility. Am J Physiol. 1993;265:F520–F529
- . Serotonergic modulation of spinal ascending activity and sacral reflex activity evoked by pelvic nerve stimulation in cats. Brain Res. 1998;798:101–108
- . Recent advances in the neurobiology of depression. Psychopharmacol Bull. 2002;36(Suppl 2):6–23
- A longitudinal evaluation of adolescent depression and adult obesity. Arch Pediatr Adolesc Med. 2003;157:739–745
Reprints not available from the authors.
Support was provided by a Grant from the National Institute of Mental Health to Dr Melville (NIMH 5 K23 MH070704). The Health and Retirement Study is funded by the National Institute on Aging (NIA) and conducted by the University of Michigan.
Cite this article as: Melville L, Fan MY, Rau H, et al. Major depression and urinary incontinence in women: temporal associations in an epidemiologic sample. Am J Obstet Gynecol 2009;201:490.e1-7.
PII: S0002-9378(09)00558-4
doi:10.1016/j.ajog.2009.05.047
© 2009 Mosby, Inc. All rights reserved.
Volume 201, Issue 5 , Pages 490.e1-490.e7, November 2009
