A Population-based Survey of the Prevalence, Potential Risk Factors, and Symptom-specific Bother of Lower Urinary Tract Symptoms in Adult Chinese Women

2.1. Study design and participants

This analysis is the second of a previous study. The data were from a national cross-sectional survey on urinary incontinence (UI) in 20 000 adult women aged ≥20 yr in China. The previous survey included two phases: a pilot study conducted in Beijing, China [11] , and a subsequent nationwide study [12] and [13]. Detailed study procedures and results for stress urinary incontinence (SUI), urgency urinary incontinence (UUI), and mixed urinary incontinence (MUI) have been described in previous reports [11], [12], and [13].

Phase 1 (a pilot survey in Beijing) assessed a representative sample of approximately 5221 respondents (from one urban district and one rural district) who were interviewed face to face by trained interviewers using a standardized questionnaire. The eligibility criteria stipulated women aged ≥20 yr in the general population who had lived in the registered areas for at least 10 yr. We excluded women who were medically diagnosed with cognitive impairment and who could not understand the questions. Detailed study procedures and results were published [11] .

Phase 2 was a nationwide survey conducted between February and July 2006 in six regions of China. The sampling process was multistage and stratified according to geographic region (northwest, southwest, north, east, northeast, and central-south China) and degree of urbanization (large cities and undeveloped economic status were assessed on the basis of the gross domestic product for each province). Eligibility and exclusion criteria were in accord with the pilot survey (on the basis of Chinese population data from 2006). The rationale for the recruitment approach, the study design, and the interpretation of results has been described previously [12] and [13]. Ethics committee approval and written informed consent were obtained.

Our previous study focused on UI, so we recalculated the sample size for this reanalysis according to a previous study conducted in Fuzhou, China, which reported 39.7% prevalence of LUTS in female residents [14] . We aimed to survey at least six age deciles: 20–29, 30–39, 40–49, 50–59, 60–69, and ≥70 yr. A minimal sample size was calculated for detection of 39.7% prevalence of LUTS, with a 5% estimated error and a 95% confidence interval (CI). Considering a 20% refusal rate, a total of 16 805 participants would be required for the LUTS reanalysis according to the formula

FORMULA:

The sample size was large enough for our reanalysis.

2.2. Diagnostic criteria

The definitions in this study are consistent with the standards recommended by the 2002 International Continence Society (ICS) guidelines [15] . Nocturia is a condition whereby a person awakens at night one or more times to void, and was defined as two or more micturitions per night in our study. Daytime frequency is a condition whereby a patient voids too often by day. In our study, frequency was defined as voiding eight times or more per day. Respondents who reported both UUI and SUI symptoms were classified as having MUI. Those who reported UI without symptoms of UUI or SUI were categorized as having other UI. Women who had given birth to multiple children were defined as having more than one delivery.

2.3. Measurements

We used a modified Bristol Female Lower Urinary Tract Symptoms questionnaire [16] that consisted of two parts. Part 1 included general information about age, marital status, occupation, menopausal status, medically diagnosed disease status, parity, mode of delivery for childbirth, and offspring birth weights. Part 2 was a Chinese version of the International Consultation on Incontinence Questionnaire–Female Lower Urinary Tract Symptoms (ICIQ-FLUTS). The Chinese version of the questionnaire was translated by survey research experts. To assess the accuracy of the translation, the Chinese version was translated back to English and compared by an independent agent. A pilot study of 40 women was performed before the survey took place to ensure the validity of the questionnaire. The readability of the Chinese version of the questionnaire was at the primary school level.

The questionnaire included an assessment of ten types of LUTS: nocturia, daytime frequency, urgency, UUI, SUI, other incontinence, pain or burning, hesitancy, straining, and intermittency. Participants were asked to rate how often individual LUTS were experienced during the previous 4 wk and, if LUTS were experienced, to what degree the symptoms were bothersome. All subjects reporting LUTS were asked to describe the frequency of the symptoms. Nocturia was measured as the number of micturitions per night, and frequency was defined as eight or more times per day. The other symptoms were measured according to how long participants had experienced them in the previous 4 wk, classified as: no (never); <10 d (occasionally); 10–20 d (sometimes); 20 d (most of the time); or every day (all of the time). The prevalence of other LUTS was defined as answers other than no. The bother associated with each symptom was evaluated using a scale ranging from 0 (not bothered at all) to 10 (greatly bothered). Bother was also defined by dividing the Likert responses into three groups: minor (rated 1–4), moderate (rated 5–7), and severe (rated 8–10). Detailed information on sociodemographic background, reproductive factors, defecation, and medical conditions was also collected. The interviewers underwent standardized training and regular quality control checks.

2.4. Statistical analysis

The data were finally stratified into eight age deciles. We performed χ² tests to compare prevalence differences between groups. Receiver operating characteristic (ROC) curve distribution was used to assess the number of deliveries and their relationship to LUTS with moderate or severe bother. A multivariate model was set up to assess bother and individual LUTS-related potential risk factors and to estimate odds ratios and 95% CIs; this model was adapted for all factors in the final version. A two-sided p value ≤0.05 was considered for statistical significance. EpiData software ( http://www.epidata.dk/ ) was used for data entry and error detection. SPSS v.12.0 (IBM Corp., Armonk, NY, USA) was used for statistical analysis.

3.1. Prevalence of LUTS

The frequencies of individual LUTS are shown in Table 2 . The prevalence of any LUTS was 55.5% and increased with age. Storage symptoms were more common than voiding symptoms (53.9% vs 12.9%). Nocturia was the most common symptom (23.4%), followed by urgency (23.3%) and SUI (18.9%); incontinence was also highly prevalent (31.9%). Straining was the least common symptom (3.3%; Table 2 ).

A notable age-related trend was observed for the prevalence of any LUTS, storage, or voiding symptoms, with approximately 50% greater prevalence observed for the oldest compared to the youngest participants ( Table 2 ; Fig 2 and Fig 3). UI was frequent in women aged ≤70 yr, but nocturia was more prevalent in those aged ≥70 yr. Urgency was usually reported by women aged ≤50 yr. Frequency, urgency, hesitancy, straining, and intermittency were more common in urban women (p < 0.001). Rural women experienced more nocturia (22.8% vs 24.1%; p = 0.032). UI was the most common symptom (28.8% vs 34.9%; p < 0.001) in both areas, but particularly in rural areas ( Fig. 3 ).

Fig. 2 Age-specific prevalence of lower urinary tract symptom subtypes. UI = urinary incontinence.

Fig. 3 Urban and rural prevalence differences for individual lower urinary tract symptoms. UI = urinary incontinence.

3.2. Symptom-specific bother of LUTS

The extent to which study participants were bothered by each urinary symptom is summarized in Table 3 and Fig. 4 . We found that no bother or minor bother were the most common responses reported for every symptom. Only 14.6–29.9% of participants with bothersome LUTS reported a moderate to severe impact on their quality of life; 4.2–11.5% reported severe bother. Some of the most prevalent symptoms were among those least commonly reported as bothersome, such as SUI (67.4%). Some of the most likely to be rated as bothersome were not necessarily the most likely to cause moderate or severe bother, such as nocturia (19.5%). Urgency was among the least likely to cause bother but was the symptom most likely to cause severe (11.5%) or moderate to severe bother (29.9%), followed by UUI (10.8% and 27.6%, respectively). Frequency was the storage symptom that was least commonly reported as causing moderate or severe bother (14.6%).

Fig. 4 Symptom-specific bother and extent of individual lower urinary tract symptoms. Bother was also defined by dividing the Likert responses into three groups: (a) 1–4, minor; (b) 5–7, moderate; and (c) 8–10, severe. UUI = urgency urinary incontinence; SUI = stress urinary incontinence.

Voiding symptoms were less likely to cause moderate or severe bother than storage symptoms (p < 0.001). Pain or burning was the second most frequent bother (84.5%), but straining and intermittency were more likely to cause moderate or severe bother (21.3% and 18.8%, respectively; Table 3 ).

3.3. Potential risk factors for LUTS

Risk factors for LUTS according to logistic regression analysis are shown in Table 4, Table 5, and Table 6. We identified multiple risk factors for any, individual, and bothersome LUTS; older age, postmenopausal status, constipation, and coexisting pelvic organ prolapse (POP) increased the odds for every type of LUTS. Older age and coexisting POP were both strong predictors (p ≤ 0.05). In comparison with spontaneous vaginal single births, our study showed a higher risk of storage and moderate or severe bothersome LUTS in women with multiple vaginal deliveries (1.46-fold and 1.26-fold, respectively) and instrumental deliveries (1.25-fold and 1.22-fold, respectively). ROC distributions showed that multiple deliveries were more closely related to any severely bothersome LUTS ( Fig. 5 ). Women who had single or multiple deliveries via cesarean section had no increase in odds for every LUTS (p > 0.05), as did women who had spontaneous vaginal single births. Participants with histories of prolonged labor and perineal laceration had increased odds for moderate or severe bothersome LUTS; prolonged labor may have a stronger influence (1.99-fold, with adjustment for all other factors). Voiding symptoms were less influenced by delivery histories. Lifestyle factors and pathologic conditions such as higher body mass index (BMI), hypertension, diabetes, alcohol consumption, and smoking had various effects on any, individual, and bothersome LUTS (Table 4, Table 5, and Table 6).

Fig. 5 Receiver operating characteristic curve analysis for the relationship between number of deliveries and moderately or severely bothersome lower urinary tract symptoms (LUTS). (A) Moderately to severely bothersome LUTS in relation to number of deliveries. Area = 0.653; 95% confidence interval 0.639–0.667. When n = 2(1.5), the sensitivity is 63.1% and the specificity is 61.3%. (B) Any severely bothersome LUTS in relation to the number of deliveries. Area = 0.689; 95% confidence interval 0.666–0.711. When n = 2(1.5), the sensitivity is 70.0% and the specificity is 60.1%.

4.1. Symptom-specific bother of individual LUTS

Assessment of the bother of symptoms is of key clinical relevance because it relates to both quality-of-life impairment and treatment-seeking [25] . Clinical practice guidelines suggest that only patients reporting bothersome symptoms should be targeted for intervention and that watchful waiting is appropriate for those with minimal symptoms [26] . There is a need to assess the frequency of individual LUTS and their respective bother ratings.

All of the individual LUTS caused bother, and storage symptoms were more likely to cause moderate or severe bother than voiding symptoms, especially urgency or UUI. We found that symptoms and bother did not correspond in a straightforward manner. Nocturia was most frequently rated as bothersome; however, it was primarily rated as no bother or as a minor bother. Urgency and UUI were less frequently rated as bothersome and usually ranked as severe or moderate. This information could assist in targeting patients who need intervention. Our study revealed similar findings to those of Moreira et al [6] , who also noted that urgency was usually associated with moderate or severe bother, whereas nocturia was usually associated with minor bother. Our study is partly in line with those of Coyne et al [27] and Møller et al [28] , who reported that the symptoms most frequently rated bothersome were nocturia, SUI, and urgency [27] and [28], and is similar to a national Finnish study that showed a high burden of urgency and UUI [29] . Because urgency seems to have a more negative impact on patient quality of life, it should be a leading consideration for medical intervention.

4.2. Potential risk factors for LUTS

The pathogenesis of LUTS is not completely understood, but it is considered a multifactorial process. A broader view of LUTS has shown that potential risk factors include age, race, micturition habits, lifestyle factors, and pathologic conditions [30] . We performed a detailed evaluation of potential risk factors for any, individual, and bothersome LUTS, and found that all were influenced by multiple factors. Older age and coexisting POP were both strong predictors for every type of LUTS. Regardless of the number of pregnancies, women who had only cesarean deliveries had no increase in odds for LUTS of any type. Women with multiple vaginal deliveries had increased odds for storage and bothersome LUTS compared with those who had spontaneous vaginal single births, in agreement with prospective studies focusing on reproductive history and progression of LUTS [31] and [32]. Prolonged labor may have a strong influence on any severely bothersome LUTS. Voiding symptoms were less likely to be associated with multiple vaginal deliveries or higher BMI. Multiple vaginal deliveries and prolonged labor may cause repeated injury to the muscles, nerves, and connective tissue of the pelvic floor, and have a negative impact on bladder control. The close relationship between LUTS and POP has rarely been explored. Having either incontinence or prolapse may be associated with an increased risk of developing the other condition [33] . It is possible that women with advanced POP experience more obstructive urinary symptoms, as reported by Rosenzweig et al [34] and Raz et al [35] . Family studies and twin studies have demonstrated that LUTS and POP are both heritable [33] . Our study, however, is cross-sectional, and the risk factors we identified cannot be considered established causes of LUTS. Some of the results should be interpreted with caution.

The strengths of our study are the large representative population-based sample of women and the high response rate. These enabled us to accurately determine the prevalence and bother associated with LUTS. The use of a professional questionnaire for urinary symptoms based strictly on criteria in the 2002 ICS guidelines also enabled us to compare our results with those for other studies.

Our study has some limitations. This analysis was the second of a study conducted 8 yr ago; the prevalence of LUTS may be higher today than at that time. Missing data were not handled well in our study. In addition, our questionnaire did not assess all LUTS, and the cross-sectional design is poor way to ascertain risk factors. Longitudinal studies are needed to better understand the temporal nature and associations with persistence of LUTS.

2.1. Study design and participants

This analysis is the second of a previous study. The data were from a national cross-sectional survey on urinary incontinence (UI) in 20 000 adult women aged ≥20 yr in China. The previous survey included two phases: a pilot study conducted in Beijing, China [11] , and a subsequent nationwide study [12] and [13]. Detailed study procedures and results for stress urinary incontinence (SUI), urgency urinary incontinence (UUI), and mixed urinary incontinence (MUI) have been described in previous reports [11], [12], and [13].

Phase 1 (a pilot survey in Beijing) assessed a representative sample of approximately 5221 respondents (from one urban district and one rural district) who were interviewed face to face by trained interviewers using a standardized questionnaire. The eligibility criteria stipulated women aged ≥20 yr in the general population who had lived in the registered areas for at least 10 yr. We excluded women who were medically diagnosed with cognitive impairment and who could not understand the questions. Detailed study procedures and results were published [11] .

Phase 2 was a nationwide survey conducted between February and July 2006 in six regions of China. The sampling process was multistage and stratified according to geographic region (northwest, southwest, north, east, northeast, and central-south China) and degree of urbanization (large cities and undeveloped economic status were assessed on the basis of the gross domestic product for each province). Eligibility and exclusion criteria were in accord with the pilot survey (on the basis of Chinese population data from 2006). The rationale for the recruitment approach, the study design, and the interpretation of results has been described previously [12] and [13]. Ethics committee approval and written informed consent were obtained.

Our previous study focused on UI, so we recalculated the sample size for this reanalysis according to a previous study conducted in Fuzhou, China, which reported 39.7% prevalence of LUTS in female residents [14] . We aimed to survey at least six age deciles: 20–29, 30–39, 40–49, 50–59, 60–69, and ≥70 yr. A minimal sample size was calculated for detection of 39.7% prevalence of LUTS, with a 5% estimated error and a 95% confidence interval (CI). Considering a 20% refusal rate, a total of 16 805 participants would be required for the LUTS reanalysis according to the formula

FORMULA:

The sample size was large enough for our reanalysis.

2.2. Diagnostic criteria

The definitions in this study are consistent with the standards recommended by the 2002 International Continence Society (ICS) guidelines [15] . Nocturia is a condition whereby a person awakens at night one or more times to void, and was defined as two or more micturitions per night in our study. Daytime frequency is a condition whereby a patient voids too often by day. In our study, frequency was defined as voiding eight times or more per day. Respondents who reported both UUI and SUI symptoms were classified as having MUI. Those who reported UI without symptoms of UUI or SUI were categorized as having other UI. Women who had given birth to multiple children were defined as having more than one delivery.

2.3. Measurements

We used a modified Bristol Female Lower Urinary Tract Symptoms questionnaire [16] that consisted of two parts. Part 1 included general information about age, marital status, occupation, menopausal status, medically diagnosed disease status, parity, mode of delivery for childbirth, and offspring birth weights. Part 2 was a Chinese version of the International Consultation on Incontinence Questionnaire–Female Lower Urinary Tract Symptoms (ICIQ-FLUTS). The Chinese version of the questionnaire was translated by survey research experts. To assess the accuracy of the translation, the Chinese version was translated back to English and compared by an independent agent. A pilot study of 40 women was performed before the survey took place to ensure the validity of the questionnaire. The readability of the Chinese version of the questionnaire was at the primary school level.

The questionnaire included an assessment of ten types of LUTS: nocturia, daytime frequency, urgency, UUI, SUI, other incontinence, pain or burning, hesitancy, straining, and intermittency. Participants were asked to rate how often individual LUTS were experienced during the previous 4 wk and, if LUTS were experienced, to what degree the symptoms were bothersome. All subjects reporting LUTS were asked to describe the frequency of the symptoms. Nocturia was measured as the number of micturitions per night, and frequency was defined as eight or more times per day. The other symptoms were measured according to how long participants had experienced them in the previous 4 wk, classified as: no (never); <10 d (occasionally); 10–20 d (sometimes); 20 d (most of the time); or every day (all of the time). The prevalence of other LUTS was defined as answers other than no. The bother associated with each symptom was evaluated using a scale ranging from 0 (not bothered at all) to 10 (greatly bothered). Bother was also defined by dividing the Likert responses into three groups: minor (rated 1–4), moderate (rated 5–7), and severe (rated 8–10). Detailed information on sociodemographic background, reproductive factors, defecation, and medical conditions was also collected. The interviewers underwent standardized training and regular quality control checks.

2.4. Statistical analysis

The data were finally stratified into eight age deciles. We performed χ² tests to compare prevalence differences between groups. Receiver operating characteristic (ROC) curve distribution was used to assess the number of deliveries and their relationship to LUTS with moderate or severe bother. A multivariate model was set up to assess bother and individual LUTS-related potential risk factors and to estimate odds ratios and 95% CIs; this model was adapted for all factors in the final version. A two-sided p value ≤0.05 was considered for statistical significance. EpiData software ( http://www.epidata.dk/ ) was used for data entry and error detection. SPSS v.12.0 (IBM Corp., Armonk, NY, USA) was used for statistical analysis.

3.1. Prevalence of LUTS

The frequencies of individual LUTS are shown in Table 2 . The prevalence of any LUTS was 55.5% and increased with age. Storage symptoms were more common than voiding symptoms (53.9% vs 12.9%). Nocturia was the most common symptom (23.4%), followed by urgency (23.3%) and SUI (18.9%); incontinence was also highly prevalent (31.9%). Straining was the least common symptom (3.3%; Table 2 ).

A notable age-related trend was observed for the prevalence of any LUTS, storage, or voiding symptoms, with approximately 50% greater prevalence observed for the oldest compared to the youngest participants ( Table 2 ; Fig 2 and Fig 3). UI was frequent in women aged ≤70 yr, but nocturia was more prevalent in those aged ≥70 yr. Urgency was usually reported by women aged ≤50 yr. Frequency, urgency, hesitancy, straining, and intermittency were more common in urban women (p < 0.001). Rural women experienced more nocturia (22.8% vs 24.1%; p = 0.032). UI was the most common symptom (28.8% vs 34.9%; p < 0.001) in both areas, but particularly in rural areas ( Fig. 3 ).

Fig. 2 Age-specific prevalence of lower urinary tract symptom subtypes. UI = urinary incontinence.

Fig. 3 Urban and rural prevalence differences for individual lower urinary tract symptoms. UI = urinary incontinence.

3.2. Symptom-specific bother of LUTS

The extent to which study participants were bothered by each urinary symptom is summarized in Table 3 and Fig. 4 . We found that no bother or minor bother were the most common responses reported for every symptom. Only 14.6–29.9% of participants with bothersome LUTS reported a moderate to severe impact on their quality of life; 4.2–11.5% reported severe bother. Some of the most prevalent symptoms were among those least commonly reported as bothersome, such as SUI (67.4%). Some of the most likely to be rated as bothersome were not necessarily the most likely to cause moderate or severe bother, such as nocturia (19.5%). Urgency was among the least likely to cause bother but was the symptom most likely to cause severe (11.5%) or moderate to severe bother (29.9%), followed by UUI (10.8% and 27.6%, respectively). Frequency was the storage symptom that was least commonly reported as causing moderate or severe bother (14.6%).

Fig. 4 Symptom-specific bother and extent of individual lower urinary tract symptoms. Bother was also defined by dividing the Likert responses into three groups: (a) 1–4, minor; (b) 5–7, moderate; and (c) 8–10, severe. UUI = urgency urinary incontinence; SUI = stress urinary incontinence.

Voiding symptoms were less likely to cause moderate or severe bother than storage symptoms (p < 0.001). Pain or burning was the second most frequent bother (84.5%), but straining and intermittency were more likely to cause moderate or severe bother (21.3% and 18.8%, respectively; Table 3 ).

3.3. Potential risk factors for LUTS

Risk factors for LUTS according to logistic regression analysis are shown in Table 4, Table 5, and Table 6. We identified multiple risk factors for any, individual, and bothersome LUTS; older age, postmenopausal status, constipation, and coexisting pelvic organ prolapse (POP) increased the odds for every type of LUTS. Older age and coexisting POP were both strong predictors (p ≤ 0.05). In comparison with spontaneous vaginal single births, our study showed a higher risk of storage and moderate or severe bothersome LUTS in women with multiple vaginal deliveries (1.46-fold and 1.26-fold, respectively) and instrumental deliveries (1.25-fold and 1.22-fold, respectively). ROC distributions showed that multiple deliveries were more closely related to any severely bothersome LUTS ( Fig. 5 ). Women who had single or multiple deliveries via cesarean section had no increase in odds for every LUTS (p > 0.05), as did women who had spontaneous vaginal single births. Participants with histories of prolonged labor and perineal laceration had increased odds for moderate or severe bothersome LUTS; prolonged labor may have a stronger influence (1.99-fold, with adjustment for all other factors). Voiding symptoms were less influenced by delivery histories. Lifestyle factors and pathologic conditions such as higher body mass index (BMI), hypertension, diabetes, alcohol consumption, and smoking had various effects on any, individual, and bothersome LUTS (Table 4, Table 5, and Table 6).