Low lean mass is associated with lower urinary tract symptoms in US men from the 2005–2006 national health and nutrition examination survey dataset

Baseline characteristics of NHANES study participants

The average age of the 959 men included in our study was 52.08 ± 7.91 years. LLM and alternate LLM (ALLM) was reported by 143 (14.91%) and 55 (5.74%) participants, respectively. Urinary hesitancy, incomplete urine emptying, urinary frequency, and nocturia was reported by 68 (7.10%), 97 (10.11%), 141 (14.70%), and 266 (27.74%) participants, respectively. Daytime LUTS and clinical LUTS (having two or more symptoms above) was reported in 26.38% (n = 253) and 13.97% (n = 134) of the study participants, respectively (Table 1, Figure 1).

Association between LLM and other clinicopathological characteristics of NHANES study subjects

Table 1 shows details of weighted demographic characteristics and other covariates of the included participants with or without LLM. We observed significantly higher prevalence of older (54.51 ± 8.87 vs. 51.72 ± 7.70 years, P = 0.0003), obese (68.54% vs. 34.77%, P < 0.0001), and current smokers (88.38% vs. 7.70%, P < 0.0001) among men with LLM compared to those without LLM. Furthermore, prevalence of urinary hesitancy (16.50% vs. 5.92%, P < 0.0001), daytime LUTS (35.44% vs. 24.09%, P = 0.0073), and clinical LUTS (27.66% vs. 11.00%, P < 0.0001) was significantly higher in men with LLM compared to those without LLM. The prevalence of incomplete urine emptying, urinary frequency, and nocturia was also higher in participants with LLM compared to those without LLM, but these differences were not statistically significant (all P > 0.05; Table 1).

Association between LUTS and other clinicopathological characteristics of NHANES study subjects

We then compared the demographic characteristics and other covariates between men with or without LUTS such as urinary hesitancy, incomplete emptying, urinary frequency and nocturia (Supplementary Table 1). The prevalence of LLM and ALLM was significantly higher among men with LUTS compared to those without LUTS. The prevalence of hypertension (50.66% vs. 37.67%, P = 0.0321), current or former smokers (P = 0.0251), LLM (28.79% vs. 11.40%, P < 0.0001), and ALLM (25.18% vs. 3.37%, P < 0.0001) was significantly higher in study subjects with urinary hesitancy compared to those without urinary hesitancy. The prevalence of ALLM was significant higher in participants with incomplete urine emptying compared to those without incomplete urine emptying (16.93% vs. 3.58%, P < 0.0001). The prevalence of LLM was not statistically significant between subjects with or without incomplete urine emptying (18.17% vs. 12.04%, P = 0.0827). Moreover, men with urinary frequency and nocturia showed higher prevalence of LLM and ALLM compared to those without urinary frequency and nocturia, but the differences were not statistically significant (all P > 0.05).

We further compared demographics of participants with or without daytime LUTS and clinical LUTS. The prevalence of hypertension (54.25% vs. 33.25%, P < 0.0001), diabetes (18.44% vs. 10.78%, P = 0.0019), sleep disorder (14.36% vs. 8.17%, P = 0.0006), current or former smokers (P = 0.0472), higher comorbidity index (P < 0.0001), LLM (17.58% vs. 10.98%, P = 0.0073) and ALLM (8.64% vs. 3.69%, P = 0.0020) was significantly higher in men with daytime LUTS than those without daytime LUTS (Supplementary Table 2). Furthermore, prevalence of hypertension (57.91% vs. 35.70%, P < 0.0001), diabetes (22.15% vs. 11.31%, P = 0.0019), current or former smokers (P = 0.0010), higher comorbidity conditions (P < 0.0001), LLM (26.72% vs. 10.55%, P < 0.0001) and ALLM (16.82% vs. 3.16%, P < 0.0001) was significantly higher among men with clinical LUTS compared to those without clinical LUTS (Supplementary Table 3).

Association between LLM and LUTS

We performed weighted multiple regression analysis to determine the association between LLM and ALLM with LUTS. The estimated odd ratios (ORs) for the relationship between LLM and LUTS are shown in Table 2 and those for the association between ALLM and LUTS are shown in Table 3.

In the non-adjusted model, urinary hesitancy (OR = 3.04, 95% CI: 1.77–5.24, P < 0.0001), daytime LUTS (OR = 1.56, 95% CI: 1.07–2.29, P = 0.0211) and clinical LUTS (OR = 2.35, 95% CI: 1.52–3.63, P = 0.0001) were significantly associated with LLM. However, the association between LLM and LUTS parameters such as incomplete urine emptying, urinary frequency, and nocturia was not statistically significant (all P > 0.05). In the adjusted model, LLM was associated with significantly higher risk of urinary hesitancy (OR = 7.76, 95% CI: 3.58–16.82, P < 0.0001), incomplete urine emptying (OR = 2.49, 95% CI: 1.28–4.85, P = 0.0070), urinary frequency (OR = 3.28, 95% CI: 1.81–5.93, P < 0.0001), daytime LUTS (OR = 3.88, 95% CI: 2.37 = 6.35, P < 0.0001) and clinical LUTS (OR = 8.11, 95% CI: 4.26–15.46, P < 0.0001), but was not associated with nocturia (OR = 1.40, 95% CI: 0.86–2.29, P = 0.1755).

Furthermore, adjusted model results showed that ALLM was significantly associated with urinary hesitancy (OR = 17.97, 95% CI: 6.88–46.90, P < 0.0001), incomplete urine emptying (OR = 4.68, 95% CI: 2.04–10.70, P = 0.0003), daytime LUTS (OR = 2.47, 95% CI: 1.20–5.08, P = 0.0136), and clinical LUTS (OR = 12.18, 95% CI: 5.26–28.22, P < 0.0001), but was not associated with urinary frequency and nocturia (all P > 0.05).

In addition, we observed significant differences in results from the non-adjusted and adjusted models for the association between LLM and LUTS parameters such as incomplete urine emptying (non-adjusted, OR = 1.34, 95% CI: 0.78–2.32; adjusted, OR = 2.49, 95% CI: 1.28–4.85), urinary frequency (non-adjusted, OR = 1.35, 95% CI: 0.85–2.16; adjusted, OR = 3.28, 95% CI: 1.81–5.93) for LLM, and the association between ALLM and LUTS (non-adjusted, OR = 1.56, 95% CI: 0.87–2.77; adjusted, OR = 2.47, 95% CI: 1.20–5.08). These differences may be due to confounding bias between lean mass loss and LUTS in the non-adjusted model, but they were eliminated in the adjusted model. Therefore, we postulate that adjusted OR is of greater clinical significance.

Data source

We obtained publicly available data from the National Health and Nutrition Examination Survey (NHANES), a cross-sectional study to assess the nutrition and health status of adults and children in the United States [36]. NHANES is conducted on a repeated two-year cycle by the National Center for Health Statistics (NCHS). All NHANES data are publicly available at https://www.cdc.gov/nchs/nhanes/.

Our study was based on the data from NHANES 2005 to 2006, since only this cycle contained information on both appendicular lean mass and questionnaires related to urinary symptoms. The NCHS Ethics Review Board granted the approval for NHANES to conduct human subject study. Written informed consent was obtained from all each participant.

Study population

Men aged ≥40 years who both answered questionnaires on LUTS and underwent body composition measurements were enrolled in our study.

Out of the originally enrolled 10348 individuals in the NHANES 2005–2006, we excluded females (n = 5268), individuals aged below 40 years old (n = 3537), individuals missing the examination data relating to LLM (n = 507), individuals missing LUTS questionnaires data (n = 63) and participants with prostate cancer (n = 14). Therefore, our final analysis included 959 men aged ≥ 40 years (Figure 1).

Body composition measurements

Body composition measurements were conducted by well-trained technicians using dual energy X-ray absorptiometry (DXA) QDR-4500 Hologic Scanner (Bedford, MA). DXA scans were not performed for participants with a self-reported use of radiographic contrast material (barium) in the seven days before scans, weight above 450 pounds and height above 192.5 cm. The data on total lean mass, appendicular lean mass (ALM), bone mineral content, bone area and fat mass were reported and available in NHANES 2005–2006. Although total lean mass was available, we could not find a proper cut-off value to define low lean mass according to previous studies regarding total lean mass data. Therefore, we used ALM to define LLM and ALLM in our analysis. ALM was defined as the sum of muscle mass of all four upper and lower extremity limbs. In our analysis, we define ALM adjusted for BMI (ALM_BMI) and ALM to classify low lean mass according to FNIH in 2014 [37]. For men, low lean mass (LLM) was defined as ALM_BMI< 0.789 and alternate low lean mass (ALLM) was defined as ALM <19.75 kg. Low lean mass and alternate low lean mass were used as exposure variables.

Lower urinary tract symptoms

LUTS were assessed based on answers to the following four questions, including: (1)“Do you usually have trouble starting to urinate?” for urinary hesitancy; (2)“After urinating, does your bladder feel empty?” for incomplete urine emptying; (3) “How often do you have urinary leakage?” for urinary frequency; and (4) “During the past 30 days, how many times per night did you most typically get up to urinate, from the time you went to bed at night until the time you got up in the morning?” for nocturia.

Nocturia was defined as waking up at least twice per night to urinate during sleep. If a participant answered yes to urinary hesitancy, or incomplete emptying, or at least 1 to urinary frequency, he was considered to have a daytime LUTS. An individual with 2 or more the mentioned symptoms above (urinary hesitancy, incomplete emptying, urinary frequency and nocturia) was considered to have clinical LUTS. Urinary hesitancy, incomplete emptying, urinary frequency, nocturia, daytime LUTS and clinical LUTS were used as outcome variables.

Covariates

The selection of covariates was mainly based on previous studies that assessed the relationship of LUTS with other exposure variables [22, 38]. We included data regarding demographic (age, race), socioeconomic conditions (education level, ratio of family income to poverty, insurance), lifestyle (alcohol intake, smoking status) and health status (body mass index, sleep disorder, diabetes, other comorbidities) as covariates.

Comorbidity index reflected the health status of patients and was of great significance for disease management, clinical treatment and prognosis [39]. Comorbid conditions included diabetes mellitus, congestive heart failure, chronic obstructive pulmonary disease (emphysema and/or chronic bronchitis), coronary artery disease, cancer and hypertension. The data about comorbid conditions was obtained from answers to the questionnaire in NHANES by the study subjects. An ordinal comorbidity index was then estimated based on the total number of combined condition reported by the study subjects as reported previously [22]. Hypertension was defined based on self-reported hypertension diagnosis, or a blood pressure measurement ≥140/90 mmHg [40]. Diabetes was defined as self-reported diabetes diagnosis, use of oral hypoglycemic agents or insulin, a fasting glucose level ≥ 126 mg/dL, or a plasma glucose level ≥ 200 mg/dL at 2 h after oral glucose tolerance test [41]. Congestive heart failure, chronic obstructive pulmonary disease, coronary artery disease and cancer were all based on a self-reported answer regarding previous diagnosis.

Categorical variables included race (Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black and other races), education level (less than high school, high school or general educational development and above high school), ratio of family income to poverty (≤1 and >1), whether have insurance (yes or not), alcohol intake (never, up to once a week, 2–3 times a week, 4–6 times a week and daily or more), body mass index (BMI in Kg/m²; categorized as <25 for normal weight, 25–29.9 as overweight and ≥30 as obese), diabetes (yes or not), smoking status (never, former or current). All detailed measurement process of these variable was publicly available at https://www.cdc.gov/nchs/nhanes/.

Statistical analysis

Statistical analysis was conducted according to CDC guidelines using Empower software (https://www.empowerstats.com/; X&Y solutions, Inc., Boston MA) and R version 3.4.3 (https://www.R-project.org, The R Foundation) [42]. All statistical analysis was performed by considering appropriate NHANES sampling weights using the complex multistage cluster survey design. Continuous variables were represented as mean with standard deviation, whereas, categorical variables were presented as frequency or percentage. Weighted student’s t test (for continuous variables) or weighted chi-square test (for categorical variables) were used to assess the differences between groups classified according to LLM and LUTS status. Weighted multiple regression analysis was performed to estimate the independent relationship between exposure variables (LLM, ALLM) and outcomes variables (urinary hesitancy, incomplete emptying, urinary frequency, nocturia, daytime LUTS, clinical LUTS) using non-adjusted model and model adjusted for age, BMI, race, education level, income, insurance, alcohol intake, diabetes, hypertension, smoking status and comorbidity index. P < 0.05 was considered statistically significant.