Objective
To describe the association between social isolation and urinary incontinence and fecal incontinence in older women.
Methods
We conducted a secondary database analysis of the National Social Life, Health and Aging Project for women aged 57 to 85 years old. Our primary outcome was self-report of often feeling isolated. We explored self-report of daily urinary incontinence and weekly fecal incontinence. Two logistic regression analyses were performed to assess the association between often feeling isolated and (1) daily urinary incontinence and (2) weekly fecal incontinence.
Results
A total of 1412 women were included in our analysis. Daily urinary incontinence was reported by 12.5% (177/1412) of community-dwelling older women. More women with daily urinary incontinence reported often feeling isolated (6.6%; 95% confidence interval [CI], 1.3−11.9 vs 2.6%; 95% CI, 1.7−3.5; P = .04) compared with women without daily urinary incontinence. Women with daily urinary incontinence had 3.0 (95% CI, 1.1−7.6) increased odds of often feeling isolated after adjusting for depressive symptoms, age, race, education, and overall health. Weekly fecal incontinence was reported by 2.9% (41/1412) of women. Weekly fecal incontinence and often feeling isolated were associated on univariable analysis (crude odds ratio, 4.6; 95% CI, 1.4−15.1). However, after adjusting for depressive symptoms, age, race, education, and overall health the association between weekly fecal incontinence and often feeling isolated was not significant (adjusted odds ratio, 0.65; 95% CI, 0.1−5.3; P = .65).
Conclusion
After adjusting for confounders, daily urinary incontinence was significantly associated with often feeling isolated. Weekly fecal incontinence was not found to be associated with often feeling isolated on multivariable logistic regression.
Urinary incontinence (UI) and fecal incontinence (FI) are common and embarrassing conditions that affect multiple aspects of a woman’s daily life and overall health.
Women with UI have been shown to have higher emotional disturbances compared with continent women. UI has been associated with loneliness, depression, and sadness. Melville et al found that women with severe UI had higher depression scores compared with continent women. In addition, similar depression scores were observed in women with severe UI compared with women who have severe medical comorbidities, such as chronic obstructive pulmonary disease and congestive heart failure. Sung et al found women with severe UI had increased depression and that depression improved with improvement in incontinence. In addition to emotional disturbances, women with UI have also been shown to have decrease physical activity level.
The multifactorial impact of FI on women’s lives is less well studied. Risk factors for FI include increasing age, female sex, increasing medical comorbidities, major depression, diabetes, and operate vaginal delivery. Women with FI were found to have a higher incidence of stroke, chronic cough, diabetes, night awakening, and falls compared with women without FI.
The impact of UI and FI on women’s social lives and personal relationships has been incompletely explored. We hypothesize that community-dwelling women with UI and FI may be less likely to participate in social activities outside their home and have less social support compared with older continent women. Our primary objective was to explore the association between social isolation in community-dwelling older women with and without UI. Our secondary objective was to explore the association of social isolation in community-dwelling older women with and without FI.
Materials and Methods
We conducted a secondary database analysis of the National Social Life, Health and Aging Project (NSHAP). The NSHAP is a cross-sectional study conducted between 2005-2006 surveying community-dwelling men and women in the United States between the ages of 57-85 years. The NSHAP was conducted to examine sexual practices and behaviors, overall health, and social networks of older adults. Adults were targeted for possible participation in the NHSAP study from a prior population-based study, the Health and Retirement Study. The NHSAP dataset is unique in that it provides adequate representation of previously understudied populations, such as Black and Hispanic women, by oversampling these targeted groups with an overall survey response rate of 75.5%. Information in the NSHAP was obtained from in home interviews by trained professional interviewers in both English and Spanish using computer-assisted personal interview (CAPI) methods. Data were collected by the National Opinion Research Center (NORC) through the University of Chicago and maintained by the Institute for Social Research (ISR) at the University of Michigan to ensure the quality of the dataset. Written exemption for this study was obtained from the Yale University Institutional Review Board as this work involved research of an existing dataset from a public source.
For this analysis, we included all women 57 to 85 years old (n = 1510). Women were excluded if they did not answer questions about or had missing data for UI and FI. Women were categorized as having daily UI if they answered “daily” to the question, “How frequently … have you had difficulty controlling your bladder, including leaking small amounts of urine, leaking when you cough or sneeze, or not being able to make it to the bathroom on time?” Women were categorized as having weekly FI if they answered “a few times a week” or “daily” to the question, “How frequently…have you lost control of your bowels (stool incontinence or anal incontinence)?” This grouping of women with daily or weekly FI was used for analysis purposes because of the small number of women with daily FI only. Recent epidemiologic studies have focused on women who report any FI in the last year. We chose to focus on women who reported more frequent symptoms of FI as we wanted to capture symptoms that would have a direct impact on women’s lives.
Our primary outcome of interest was the self-report of often feeling isolated in response to the question, “How often do you feel isolated from others?” Social isolation was measured through quantifying perceived social isolation, social disconnectedness, and loneliness. Cornwell and Waite explored 17 indicators of social isolation to create 2 distinct scales: perceived social isolation and social disconnectedness. Perceived social isolation was measured with questions discussing the frequency of opening up to family, frequency of relying on family, frequency of opening up to friends, frequency of relying on friends, frequency of opening up to spouse/partner, and frequency of relying on spouse/partner with 3 answer options ranging from “hardly ever or never” to “often.” Social disconnectedness was quantified by evaluating social network size, social network range, frequency of interaction, proportion of social network members in the home, number of friends, frequency of attending group meetings, frequency of socializing with friends and family, and frequency of volunteering. For this study, we looked at women with 4 or more friends, socializing with family and friends more than weekly, volunteering in the past year, and attending a group meeting in the past year. Finally, a modified version of the University of California, Los Angeles (UCLA) loneliness scale was to quantify symptoms of loneliness. The UCLA loneliness scale was first developed by Russell et al in 1978. The shortened version used by the NSHAP was developed and validated in 2004 using 3 questions on a 3-point scale. In addition to feelings of isolation, respondents were asked how often they felt a lack of companionship and how often they felt left out. Modified UCLA loneliness scale scores ranged from 3 to 9, with higher scores indicating greater feelings of loneliness.
Demographics including age, race, education, and self-reported overall health were recorded. We also reported scores from the modified Center for Epidemiological Studies-Depression (CES-D) Scale and Hospital Anxiety and Depression Scale (HADS) as to quantify depressive symptoms and symptoms of anxiety. The modified CES-D scale is an 11-question screening test for depressive symptoms. Each question had a score from 0-3, with score scales ranging from 0−31 and higher scores indicating more depressive symptoms. The HADS was developed by Zigmond and Snaith in 1983 with 7 questions relating to anxiety in hospital patients. Each question was graded by a Likert scale of 0-3, with scores ranging from 0−21 and higher scores reflecting greater symptoms of anxiety.
Statistical analyses, including descriptive and inferential statistics with percentage estimates and 95% confidence intervals (CIs), were performed as appropriate. The NSHAP dataset allowed data to be weighted to provide an estimate of population characteristics representative of community-dwelling older Americans aged 57-85 years. Survey weights were used to account for the differential probability of inclusion in the sample and differential probability of nonresponse in all the analyses. Percentage estimates and 95% CIs were obtained to report weighted frequencies. Model fitting and variance estimates used in the construction of CI account for the stratified and clustered nature of the design to produce unbiased estimates of standard errors.
Two logistic regression analyses were performed to assess the association between self-report of often feeling isolated (primary outcome) and (1) daily UI and (2) weekly FI. Potential confounders were considered for inclusion in the final models based on their significance in univariable analysis ( P ≤ .1). Significant confounders were included in the final adjusted models if they continued to impact the final model. Statistical analyses were performed using SAS 9.2 (SAS Institute, Inc, Cary, NC) and STATA 11.0 (StataCorp, College Station, TX).
Results
A total of 1412 women were included in our analysis. Seventy-six women and 22 women were excluded because of missing data in regard to UI and FI, respectively.
Daily UI was reported by 12.5% (177/1412) of community-dwelling older women. Weekly, monthly, or yearly UI was reported by 40.2% (568/1412) of women. Forty-seven percent (667/1412) of women reported no urinary incontinence in the past year.
Weekly or daily FI was reported by 2.9% (41/1412) of women with 13 women reporting daily FI. Ninety-two percent (1259/1412) of women reported no FI in the past 12 months and 8.2% (112/1412) of women reported monthly or yearly FI. Symptoms of both daily UI and weekly FI were reported by 1.2 % (17/1412) of women.
UI
Women with daily UI were older than women without daily UI (70 years ± 0.7 vs 68 years ± 9.3; P = .01) ( Table 1 ). No differences in race or education level were observed. Overall health reported as “excellent” or “very good” in 25.1% (44/177) of women with daily UI and 43.4% (534/1235) of women without daily UI ( P < .01). Women with daily UI had significantly more depressive symptoms than women with no daily UI as measured by a higher mean CES-D score (6.7 ± 0.5 vs 5.5 ± 0.2; P = .02). Women with daily UI had more anxiety symptoms measured by higher mean HADS scores compared with women with no daily UI (5.2 ± 0.5 vs 3.7 ± 0.1; P < .01).
| Variable | Daily UI | No daily UI a | P value |
|---|---|---|---|
| n = 177 | n = 1235 | ||
| Age (y, mean ± SEM) | 70 ± 0.7 | 68 ± 0.3 | .01 |
| Age category | .03 | ||
| 57-64 | 36.7 (28.1–45.3) | 40.7 (37.4–44) | |
| 65-74 | 29 (22.1–35.8) | 35.5 (32.4–38.7) | |
| 75-85 | 34.4 (25.2–43.5) | 23.8 (21.3–26.4) | |
| Race/ethnicity | .32 | ||
| White | 84.5 (76.8–92.3) | 80.7 (76.9–84.5) | |
| Black | 8.3 (4.3–12.4) | 11.1 (8.2–14) | |
| Hispanic, nonblack | 6.2 (0–12.9) | 6.1 (3.4–8.8) | |
| Other | 0.9 (0–2.2) | 2.2 (0.9–3.4) | |
| Education | .83 | ||
| Less than high school | 19.9 (12.2–27.5) | 18.2 (15.3–21.2) | |
| High school or equivalent | 32.2 (24–40.5) | 29.5 (26.4–32.7) | |
| Some college | 30.2 (21.4–39.1) | 33.3 (29.7–36.9) | |
| Bachelor’s degree or higher | 17.6 (9.5–25.8) | 18.9 (15.6–22.1) | |
| Self-reported health | < .01 | ||
| Poor | 8 (3–13) | 6 (4.4–7.5) | |
| Fair | 26.1 (18–34.1) | 15.9 (13.6–18.2) | |
| Good | 38 (29–46.9) | 30.6 (27.3–33.8) | |
| Very good | 19.8 (12.1–27.6) | 33.2 (30.4–36) | |
| Excellent | 8.1 (3–13.2) | 14.3 (11.2–17.5) | |
| How happy are you? | < .01 | ||
| Extremely happy | 5.3 (2–8.7) | 14.7 (12–17.4) | |
| Very happy | 38.3 (30.6–46) | 40.4 (37.4–43.4) | |
| Pretty happy | 44 (35.5–52.5) | 35.4 (32.6–38.2) | |
| Unhappy | 8.9 (7.4–10.4) | ||
| Sometimes | 10.7 (4.5–17) | 0.6 (0.2–1) | |
| Unhappy usually | 1.7 (0–3.4) | ||
| CES-D Score (mean ± SEM) | 6.7 ± 0.5 | 5.5 ± 0.2 | .02 |
| HADS anxiety index (mean ± SEM) | 5.2 ± 0.5 | 3.7 ± 0.1 | < .01 |
a “No Daily UI” includes women with no UI in the past 12 mo (n = 667) and women with weekly, monthly, or yearly UI (n = 568).
Perceived isolation
More women with daily UI reported often feeling isolated (6.6%, 95% CI, 1.3−11.9 vs 2.6%, 95% CI, 1.7−3.5; P = .04), often feeling a lack of companionship (12.2%, 95% CI, 5.1−19.2 vs 5.3%, 95% CI, 3.8−6.9; P < .01), and often feeling left out (7.95%, 95% CI, 2.3−13.5 vs 2.7%, 95% CI, 1.7−3.6; P < .01) compared with women without daily UI ( Table 2 ). However, women with daily UI were similar to women without daily UI in their reports of often opening up to family, often relying on family, often opening up to friends/spouse/partner and often relying on spouse/partner.
| Variable | Daily UI | No daily UI | P value |
|---|---|---|---|
| n = 177 | n = 1235 | ||
| Often feel isolated | 6.6 (1.3–11.9) | 2.6 (1.7–3.5) | .04 |
| Perceived social isolation | |||
| Often feel left out | 7.9 (2.3–13.5) | 2.7 (1.7–3.6) | < .01 |
| Often feel that you lack companionship | 12.2 (5.1–19.2) | 5.3 (3.8–6.9) | < .01 |
| Often open up to family | 59.5 (51.2–67.8) | 56.2 (53.2–59.2) | .41 |
| Often rely on family | 72.9 (65.6–80.3) | 71.8 (68.2–75.3) | .76 |
| Often open up to friends | 35.9 (26.6–45.2) | 35.6 (31.9–39.4) | .96 |
| Often rely on friends | 55 (45.2–64.8) | 50.7 (46.5–54.8) | .39 |
| Often open up to spouse/partner | 74.2 (63.7–84.8) | 76.5 (72.8–80.2) | .69 |
| Often rely on spouse/partner | 85.2 (77.1–93.3) | 83.9 (81–86.8) | .78 |
| Social disconnectedness | |||
| Number of friends ≥4 | 68.2 (58.6–77.7) | 77.9 (74.7–81.1) | .04 |
| Socializing weekly with friends | 15 (8.4–21.6) | 19.1 (16.1–22) | .27 |
| Volunteer in past year | 62.5 (51.2–73.8) | 68.5 (64.9–72.1) | .29 |
| Attend group meeting in past year | 68.3 (57.6–78.9) | 72.5 (68.9–76.1) | .41 |
| UCLA loneliness scale (mean ± SD) | 4.5 ± 0.2 | 4.0 ± 0.1 | < .01 |
When we excluded women who reported weekly, monthly, and yearly UI from the analysis, we had similar findings in regard to the primary outcome, often feeling isolated.
Social disconnectedness
Socializing with friends and family more than weekly, volunteering in the past year, and attendance at group events in the past year were similar between women with daily UI and women without daily UI ( Table 2 ). However, less women with daily UI reported having more than 4 friends compared with women without daily UI (68.2%, 95% CI, 58.6−77.7 vs 77.9%, 95% CI, 74.7−81.1; P = .04).
Loneliness
Mean score of the UCLA loneliness scale was 4.5 ± 0.2 for women with daily UI and 4.0 ± 0.1 for women without daily UI ( P = .003).
Regression analysis
Women with daily UI had 3.0 (95% CI, 1.1−7.6) increased odds of often feeling isolated after adjusting for CES-D score, age category, race, education level, and self-reported overall health ( Table 3 ).
| Variable | Adjusted OR | 95% CI | P value |
|---|---|---|---|
| Daily UI | 3.0 | 1.1–7.6 | .03 |
| CES-D score | 1.3 | 1.2–1.4 | < .01 |
| Age group, y (57-64 = referent) | .03 | ||
| 65-74 | 0.2 | 0.1–0.9 | |
| 75-85 | 1.7 | 0.8–3.8 | |
| Race (white = referent) | < .01 | ||
| Black | 4.6 | 1.9–11.1 | |
| Hispanic, nonblack, other | 0.9 | 0.1–6.5 | |
| Education (less than high school = referent) | .29 | ||
| High school or equivalent | 0.7 | 0.2–2.8 | |
| Some college | 0.6 | 0.2–2.0 | |
| Bachelor’s degree or higher | 0.2 | 0–1.2 | |
| Self-reported health (poor = referent) | .23 | ||
| Fair | 0.6 | 0.2–1.8 | |
| Good | 1.1 | 0.4–2.8 | |
| Very good | 0.2 | 0.1–0.9 | |
| Excellent | 1.3 | 0.3–6.3 |
a Modeling event = Often feel isolated (vs hardly ever, never, some of the time).
FI
Women with weekly FI were more likely to report less than high school education compared with women without weekly FI (45.8%, 95% CI, 25−66.7 vs 17.6%, 95% CI, 14.4−20.8; P = .02) ( Table 4 ). Women with weekly FI were also likely to report “poor” health compared with women without weekly FI (33.0%, 95% CI, 14.9−51.2 vs 5.4%, 95% CI, 4−6.9; P < .01). Women with weekly FI had more depressive symptoms measured by higher mean CES-D scores compared with women without weekly FI (9.5 ± 1.2 vs 5.5 ± 0.1; P < .01). In addition, they had more anxiety symptoms measured by higher mean HADS scores FI (6.3 ± 1 vs 3.8 ± 0.1; P < .01).
| Variable | Weekly FI | No weekly FI a | P value |
|---|---|---|---|
| n = 41 | n = 1371 | ||
| Age (y, mean ± SEM) | 68.5 ± 1.5 | 68.2 ± 0.3 | .87 |
| Age category | .72 | ||
| 57-64 | 39.8 (17.9–61.7) | 40.2 (36.9–43.4) | |
| 65-74 | 39.8 (21.4–58.2) | 34.6 (31.6–37.6) | |
| 75-85 | 20.4 (6–34.8) | 25.2 (22.6–27.8) | |
| Race/ethnicity | .90 | ||
| White | 82 (70.2–93.8) | 81.1 (77.2–85) | |
| Black | 10.1 (2.3–17.9) | 10.8 (7.9–13.7) | |
| Hispanic, nonblack | 6.9 (0–14.5) | 6.1 (3.1–9.1) | |
| Other | 1 (0–3.1) | 2 (0.9–3.2) | |
| Education | .02 | ||
| Less than high school | 45.8 (25–66.7) | 17.6 (14.4–20.8) | |
| High school or equivalent | 23.8 (8.9–38.7) | 30.1 (27.1–33) | |
| Some college | 16 (3.5–28.6) | 33.5 (29.7–37.3) | |
| Bachelor’s degree or higher | 14.3 (0–31.8) | 18.9 (15.8–22) | |
| Self-reported health | < .01 | ||
| Poor | 33 (14.9–51.2) | 5.4 (4–6.9) | |
| Fair | 27.2 (13.2–41.2) | 16.8 (14.6–19.1) | |
| Good | 14.3 (2.8–25.8) | 32 (29–34.9) | |
| Very good | 15.1 (0–33) | 32.1 (29.5–34.8) | |
| Excellent | 10.4 (0–22.8) | 13.7 (10.8–16.6) | |
| How happy are you? | .24 | ||
| Extremely happy | 2 (0–6.3) | 0.7 (0.3–1.1) | |
| Very happy | 5.1 (0–10.8) | 9.3 (7.9–10.7) | |
| Pretty happy | 49.6 (29.9–69.4) | 36 (33.3–38.7) | |
| Unhappy sometimes | 30.3 (12.8–47.9) | 40.4 (37.6–43.3) | |
| Unhappy usually | 13 (0–31.3) | 13.6 (11.3–15.9) | |
| CES-D score (mean ± SEM) | 9.5 ± 1.2 | 5.5 ± 0.1 | < .01 |
| HADS anxiety index (mean ± SEM) | 6.3 ± 1 | 3.8 ± 0.1 | < .01 |
| UCLA loneliness scale (mean ± SEM) | 6.1 ± 0.5 | 5 ± 0.1 | .03 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
