Changes in Bladder Health over Time: A Longitudinal Analysis of Adult Women in the Boston Area Community Health Survey.

The BACH Survey is a population-based, longitudinal study of Boston residents. Participants were sampled randomly using a two-stage cluster design, with stratification by sex, race/ethnicity (equally distributed across Black, Hispanic, and White), and age (30-39, 40-49, 50-59, and 60-79 years). 5,506 (3,205 female and 2,301 male) of 8,702 eligible participants completed the baseline visit from 2002-5. This visit included an in-home interview, self-administered questionnaire, and measurement of height, weight, and waist circumference. 8 Visit procedures were repeated at the follow-up visit approximately five years later (2008-10). We limited the present analysis to female participants who completed the baseline and follow-up interviews and provided information on at least one LUTS and one interference item at baseline and follow-up. BACH was approved by the New England Research Institutes Institutional Review Board and all participants provided written informed consent. At baseline and follow-up, participants completed questions on several aspects of bladder dysfunction, including storage (e.g., urgency incontinence, nocturia, bladder-filling pain), emptying (e.g., hesitancy, incomplete emptying, and urination-related pain), and bioregulatory (e.g., urinary tract infections [UTIs]) dysfunction. As most of these aspects of bladder dysfunction are symptoms, we refer to them hereafter as “LUTS” for simplicity. LUTS were assessed using the American Urological Association Symptom Index, 9 Sandvik Incontinence Severity Scale, 10 Interstitial Cystitis Symptom Index, 11 and items written specifically for BACH. Items with ≤20% disagreement across responses were combined to create 15 distinct LUTS, using the maximum response from any of the contributing items to capture LUTS frequency ( Appendix Table 1 ). Bladder function-related well-being was assessed in all women irrespective of whether or not they endorsed LUTS by the interference with activities domain of the Epstein Quality of Life Questionnaire for benign prostatic hyperplasia. 12 This validated scale, which was administered at both baseline and follow-up, queries the frequency of interference from urinary problems in the past month with seven activities. It was modified for BACH to refer to urinary “problems” as “experiences” and to include interference due to pain or discomfort in the pubic area to increase its relevance for interstitial cystitis/bladder pain syndrome. 13 The Epstein scale has been used extensively for participants of both sexes in BACH. 13 - 16 For all analyses, we weighted observations inversely proportional to their probability of selection into BACH, with further post-stratification to the Boston population using the 2000 US Census. 8 To investigate changes in bladder health between the baseline and follow-up interviews, we used data on the maximum frequency of LUTS (across 15 symptoms) and interference (across seven activities) to create five bladder health categories at baseline and follow-up: 1) no LUTS or interference (i.e., optimal bladder health); 2) mild LUTS (maximum of “rarely”) or interference (maximum of “a little”); 3) moderate LUTS (“a few times”) or interference (“some of the time”); 4) severe LUTS (“fairly often” or “usually”) or interference (“most of the time”); and 5) very severe LUTS (“almost always”) or interference (“all of the time”). 6 , 7 Women who reported past or current bladder conditions/treatment were included in the highest category irrespective of their symptoms. We also explored changes in bladder health over time by calculating a continuous bladder health score at baseline and follow-up. This score was calculated as the sum of standardized scores for LUTS (sum of the frequency of 12 LUTS [scales from 0 to 4/5]) and interference (sum of the frequency of interference with seven activities [scales of 0-4] 12 ), for a combined score ranging from 0 (best) to 56 (worst bladder health). 6 To investigate whether baseline bladder health is associated with changes in bladder health over time, we used ordinal logistic regression and generalized linear models to calculate crude- and multivariable-adjusted odds ratios (ORs) and relative risks (RRs), respectively. We also performed stratified analyses by age group and sensitivity analyses excluding women with: 1) known past or current bladder conditions/treatment (current LUTS medications, previous incontinence or bladder surgery, chronic indwelling catheterization, and bladder cancer), as their exact position on the bladder health distribution was assumed, not known based on their untreated LUTS/well-being; and 2) non-bladder conditions that might contribute to LUTS or interference (genitourinary cancers besides bladder cancer, prolapse of the uterus, “bladder, or rectum”, congenital urinary tract abnormalities [many of which are renal 17 , 18 ], endometriosis, pelvic inflammatory disease, chronic pelvic pain, vulvodynia, and diabetes). Analyses were performed using R v3.2.4 and SAS® 9.3.

Of the 3,205 female participants interviewed at baseline, 2,534 were interviewed approximately 5 years later, and 2,526 provided information on LUTS and interference at both time points. The majority of included participants were young to middle-aged, parous, former or current smokers, and overweight or obese ( Appendix Table 2 ). Compared to excluded participants, included participants were more likely to be White, peri-menopausal, former or current smokers, and consumers of 1-2 alcoholic beverages/day, with greater health-related limitations in activities, and a lesser likelihood of comorbid conditions. They also reported a slightly higher prevalence of LUTS, but a lower prevalence of interference with activities. At baseline, a similar distribution of bladder health was observed as in previous analyses: 16.8% reported optimal bladder health (no LUTS or interference), 61.0% reported intermediate health (15.3% mild LUTS/interference, 21.9% moderate LUTS/interference, and 23.8% severe LUTS/interference), and 22.3% reported poor bladder health (LUTS/interference all the time; Figure 1 ). When bladder health was examined as a continuous measure, the distribution was skewed to the right with a median value of 2.4 (interquartile range [IQR]=0.5-7.7 out of 0-56). Considering changes in bladder health over time, only a small percentage of women (6.5%) reported optimal bladder health at baseline and follow-up ( Figure 2 ). Approximately one third (33.6%) developed worse bladder health, including 10.4% who transitioned from optimal to less than optimal bladder health. This reflects an incidence of 61.5% worsened bladder health among those with optimal bladder health at baseline (20.1% to mild LUTS/interference, 18.3% to moderate LUTS/interference, 13.0% to severe LUTS/interference, and 10.1% to very severe LUTS/interference). In contrast, 28.7% of participants improved in bladder health, including 11.0% who resolved their LUTS or interference completely; these values are 34.5% and 13.2%, respectively, when limited to women with LUTS/interference at baseline rather than all participants. A further 31.4% of participants maintained the same bladder health category as at baseline. Combining all women, approximately one third (37.9%) remained in the same bladder health category over time, one third (33.6%) worsened, and approximately one third improved (28.7%), resulting in generally similar distributions of bladder health at baseline and follow-up. These distributions shifted slightly more towards worsening with older age ( Appendix Table 3 ) and slightly more towards changes in bladder health (improvement and worsening) when women with current or past bladder conditions and those that might influence LUTS or interference were excluded from the analyses (data not shown). When we modeled bladder health as a continuous measure, the mean change was 0.7 (standard deviation=8.0) over follow-up ( Appendix Figure 1 ), with a correlation between baseline and follow-up bladder health of 0.50. Changes in bladder health, including worsening, maintenance, and improvement, were observed in almost all categories of bladder health ( Figure 2 ). However, women with better bladder health at baseline were more likely to worsen (the only direction of change possible for many of them) and women with poorer bladder health at baseline were more likely to improve. Overall, these patterns resulted in inverse associations between poorer baseline bladder health categories and risks of worsening or worsening and maintained bladder health over time (e.g., multivariable-adjusted RR=0.55, 95% confidence interval [CI]: 0.49-0.63 for worsening or maintained LUTS/interference in women with very severe LUTS/interference at baseline, Table 1 ). Nevertheless, women in a poorer bladder health category at baseline were still more likely to be in a poorer bladder health category at follow-up than women in better categories (e.g., multivariable-adjusted OR=15.2, 95% CI: 9.55-24.3 per higher LUTS/interference category at follow-up in women with very severe LUTS/interference at baseline). They were also more likely to have severe or very severe LUTS or interference at follow-up (e.g., RR=3.27, 95% CI: 2.49-4.29 in women with very severe LUTS/interference at baseline). Similar patterns of associations were observed when LUTS and interference were examined separately, and when women with a history of bladder conditions or treatment were excluded from the analyses (data not shown).

In this large, regionally-representative analysis of bladder health in American women, we observed considerable variability in bladder health over time. Only a small percentage of women (6.5%) had optimal bladder health over five years, although a considerable percentage improved over time (28.7%). The remaining women either maintained their bladder health status (31.4%), or worsened (33.6%), including 10.4% who transitioned from optimal to less than optimal health. Worsening, maintenance, and improvement were observed in almost all categories of bladder health, although women with LUTS/interference at baseline were more likely to have LUTS/interference at the 5-year follow-up interview and to be in a higher LUTS/interference category, including severe or very severe LUTS/interference, than women without LUTS/interference at baseline. To our knowledge, our study is one of the first to examine the incidence of worsening bladder health and changes in bladder health over time (assessed by both LUTS and interference), precluding direct comparisons to findings from previous studies. However, several prior studies have examined LUTS incidence or changes in LUTS over time, 19 - 27 allowing us to make, at least, partial comparisons. Despite differences in LUTS examined, definitions of LUTS and LUTS change used, frequency of LUTS assessment, and length of follow-up time, our findings of considerable variability over time are similar to those from previous studies, most of which observed LUTS incidence and/or progression over time (0.4-37.5%), as well as some degree of regression or resolution (7.8-56.0% improved among those with LUTS at baseline). 19 - 27 Notably, many of these studies also observed changes in LUTS in almost all baseline LUTS categories. 19 - 22 , 24 While the development or progression of LUTS has been studied extensively, remission or resolution of LUTS or, in our case, improvement in bladder health, is much less well-studied. Several potential factors may explain this phenomenon. The first of these is treatment. Although we grouped women who underwent LUTS surgery or used LUTS medications in the highest bladder health category and did not allow them to change categories over time, it is possible that other therapies or management strategies not captured on BACH questionnaires (e.g., vaginal estrogen, biofeedback, and medications for diabetes, anxiety, and depression) may have contributed to improvement in bladder health over time. Changes in some conditions or states (e.g., pregnancy, menopausal transition) may have also resulted in remission of LUTS and/or interference. A second possible explanation is adaptation to LUTS, leading to reduced interference. Possible adaptations include avoiding factors that trigger LUTS, using absorptive incontinence products, modifying fluid intake and type, discontinuing certain activities, and accepting LUTS as a “normal” part of aging. A third possibility is natural regression or resolution of LUTS and/or interference and regression to the mean, and a fourth is behavioral modification to reduce LUTS and interference. Currently, only a few such options are known (e.g., weight loss and pelvic floor muscle exercises), but this possibility is encouraging and should be explored further to inform secondary prevention strategies. Although we observed that some women in all categories of baseline bladder health improved over time, women in poorer categories of bladder health at baseline were significantly more likely to be in poorer categories of bladder health five years later than women in initially better categories. This same association has also been observed for LUTS in previous studies. Specifically, women with a greater frequency of urinary incontinence at baseline were more likely to have persistent or worsening urinary incontinence over time than women with lower frequencies of or without urinary incontinence in the few previous studies that have examined this association to date. 21 , 25 , 28 Together, these findings underscore the importance of maintaining good bladder health and support the need for early interventions to promote good bladder health and prevent bladder health deterioration over time (i.e., primary prevention strategies). Our analysis takes advantage of several strengths of the BACH Survey to investigate changes in bladder health over time. These include its large sample size; representative sampling of community-dwelling women; racial/ethnic diversity; extensive collection of information on LUTS and bladder conditions/therapies; rare collection of bladder function-related well-being on all women irrespective of LUTS; and five-year follow-up. Limitations of our analysis include its use of a crude measure of bladder health that does not take into account other elements of bladder health, such as emotional aspects of bladder function-related well-being and ability to adapt to short-term stressors; reliance on self-reported LUTS and interference data; and lack of information on LUTS, interference, and factors that influence these two variables in the five years between interviews. In summary, our findings demonstrate considerable variability in bladder health over time, with the potential for worsening, maintenance, and improvement, even in the absence of LUTS therapy. Together with our observed positive associations between poorer bladder health categories at baseline and follow-up, these findings suggest the need for additional primary and secondary prevention efforts to promote good bladder health and prevent the initial onset and progression of poor bladder health.

Health, as defined by the World Health Organization, is “a state of complete physical, mental, and social well-being and not merely the absence of disease and infirmity.” 1 This definition has been extended to several individual organs, including the heart, 2 brain, 3 and, most recently, the bladder. 4 Specifically, members of the Prevention of Lower Urinary Tract Symptom Research Consortium developed the following research bladder health definition: “A complete state of physical, mental, and social well-being related to bladder function and not merely the absence of lower urinary tract symptoms (LUTS).” This state “permits daily activities, adapts to short-term physical or environmental stressors, and allows optimal well-being (e.g., travel, exercise, social, occupational, or other activities).” 4 The Consortium also applied this definition to the three main functions of the bladder (“storage,” “emptying,” and “bioregulatory”) and described unhealthy characteristics of each function. 5 To begin to inform and quantify the spectrum of bladder health, we recently took advantage of data from the Boston Area Community Health (BACH) Survey to estimate a crude distribution of bladder health in community-based women. Using information on LUTS and bladder function-related well-being (measured by interference with activities from urinary experiences), we found that approximately one in five women might be considered to have optimal bladder health (no LUTS or interference), three in five to have intermediate health (intermediate frequencies of LUTS/interference), and one in five to have poor bladder health (LUTS/interference all the time). 6 This distribution was essentially unchanged when we used number rather than frequency of LUTS and additional measures of bladder function-related well-being, including life impact, bladder-related thought, and perception of bladder condition. 7 Together, these findings provide preliminary insight into the spectrum of bladder health and suggest wide variability across women. However, no studies, to our knowledge, have examined variability in bladder health over time. Therefore, we used data from the baseline and 5-year follow-up interviews of BACH to describe changes over time.