Abstract
Objective
Waist circumference (WC) has been linked to exacerbations in chronic obstructive pulmonary disease (COPD), but the relationship between WC, height, and the Body Roundness Index (BRI) with mortality in COPD patients remains unclear. This study investigates the association between BRI and both all-cause and cardiovascular mortality in COPD patients.
Methods
Data from 3,672 COPD participants (1,517 men and 2,155 women; mean age: 53.17 ± 16.46 years) were obtained from NHANES (1999–2018). Multivariable Cox proportional hazards models, Kaplan–Meier analysis, Nelson-Aalen cumulative hazard plots, and restricted cubic spline analyses were used to assess the relationship between BRI and mortality. ROC curves were constructed to evaluate BRI’s predictive performance. Subgroup and sensitivity analyses ensured robustness of the models.
Results
The BRI was associated with all-cause (HR: 1.12, 95% CI: 1.05–1.19) and cardiovascular mortality (HR: 1.14, 95% CI: 1.01–1.29). Kaplan–Meier curves and Nelson-Aalen plots demonstrated that higher BRI quartiles correlated with lower survival probabilities and increased cumulative mortality incidence. ROC curve analysis showed that BRI outperformed other models in predicting all-cause [AUC: 0.81 (0.79–0.82)] and cardiovascular mortality [AUC: 0.79 (0.76–0.81)]. A significant interaction was observed between BRI and alcohol consumption for all-cause mortality. Results remained consistent after excluding participants who died within two years (HR: 1.10, 95% CI: 1.03–1.18) and in the alcohol-consuming subgroup (HR: 1.14, 95% CI: 1.05–1.24).
Conclusion
The BRI is associated with increased all-cause and cardiovascular mortality in COPD patients, with higher BRI quartiles linked to greater mortality risk. Targeting BRI may offer a potential strategy to reduce mortality in this population.
Keywords: BRI, COPD, All-cause mortality, Cardiovascular mortality
Background
Chronic obstructive pulmonary disease (COPD) is a prevalent respiratory condition characterized by persistent airflow limitation and progressive disease progression. If untreated, COPD can impact other systems, including the cardiovascular, hepatic, renal, and central nervous systems, resulting in associated complications and a high rate of morbidity and mortality [1, 2]. COPD is currently the third leading cause of death worldwide [3, 4]. Currently, there is no cure for COPD, and extending the survival of COPD patients remains a primary therapeutic goal.
In 2013, Thomas et al. introduced the Body Roundness Index (BRI) as a tool to estimate body fat and visceral fat volume [5]. Compared to traditional body composition indicators, such as waist circumference (WC), BMI, and hip circumference, the BRI more accurately reflects body fat percentage and visceral fat distribution. Previous studies on obesity have predominantly relied on BMI; however, BMI is heavily influenced by height and does not accurately predict fat distribution. It cannot differentiate between muscle mass and fat tissue, nor can it reflect visceral fat [6]. Furthermore, studies have shown that body weight contributes minimally to the prediction of pulmonary function parameters [7–9]. Vague et al. proposed classifying obesity into central obesity (visceral obesity) and peripheral obesity. Central obesity refers to fat primarily distributed in the abdominal viscera and is more strongly associated with disease occurrence compared to other forms of obesity [10]. Research by Vgontzas et al. demonstrated a positive correlation between the respiratory disturbance index and visceral fat, with no association to BMI, total fat, or subcutaneous fat [11]. Recent studies using CT and MRI to measure abdominal subcutaneous and visceral fat have shown that waist circumference and waist-to-hip ratio are reliable, simple indicators of central obesity, with waist circumference being the more commonly used [12]. Therefore, the BRI is an effective measure for identifying central obesity and is associated with insulin resistance [5]. Studies have shown that the BRI is a potential risk factor for coronary artery disease and carotid atherosclerosis, with elevated BRI associated with an increased risk of cardiovascular disease [13]. Research by Li et al. found that BRI is associated with metabolic syndrome, insulin resistance, and inflammatory markers, and is the most effective measure for identifying insulin resistance in obese and overweight populations [14]. The BRI also offers an advantage over BMI and waist circumference in assessing the risk of metabolic diseases across different gender groups, with greater gender differentiation than BMI and WC [15].
Recent studies have shown that an increase in abdominal fat is significantly correlated with indicators of COPD severity, such as the Apnea–Hypopnea Index (AHI) and minimum oxygen saturation [16, 17]. While numerous studies have confirmed the correlation between waist circumference, waist-to-hip ratio, and hypoxic conditions, research on the relationship between BRI and COPD remains limited, and large-scale, multicenter studies are lacking. In conclusion, COPD has a high mortality rate, and central obesity is closely related to the severity of COPD. Therefore, in order to clarify the relevant risk factors and reduce the mortality rate of COPD, this study explored the correlation between BRI index and all-cause mortality and cardiovascular mortality in COPD population.
Materials and methods
Data source
The data used in this study were obtained from the National Health and Nutrition Examination Survey (NHANES) database, a cross-sectional survey of U.S. residents. NHANES systematically collects health-related data from a nationally representative sample of U.S. residents using a stratified, multistage probability sampling design [18]. Data are collected through household interviews and standardized physical examinations conducted at specially equipped mobile examination centers. NHANES is a collaborative effort between the Centers for Disease Control and Prevention (CDC) and the National Center for Health Statistics (NCHS). The NHANES protocol was approved by the NCHS Research Ethics Review Board, and written informed consent was obtained from each participant (source: https://www.cdc.gov/nchs/nhanes/index.htm). Ethic declarations: Human Ethics and Consent to Participate declarations: not applicable.
Population selection
To evaluate the association between BRI-related indices and all-cause and cardiovascular mortality in the COPD population, only individuals diagnosed with COPD were included. A total of 101,316 individuals were examined across ten interview cycles from 1999 to 2018. After excluding individuals without COPD, 4,194 participants were initially retained. Further exclusions were made for individuals without waist circumference (WC) or height measurements, or those with missing follow-up data. Ultimately, 3,672 participants were retained for this study (Fig. 1).
Fig. 1.
Flowchart of participant inclusion
COPD diagnosis
COPD diagnosis was confirmed based on responses to the Medical Conditions Questionnaire (MCQ), which included the following questions:"Has a doctor ever told you that you have chronic obstructive pulmonary disease?"(MCQ160o),"Has a doctor ever told you that you have emphysema?"(MCQ160g), and"Has a doctor ever told you that you have chronic bronchitis?"(MCQ160k) [19]. Participants who answered affirmatively to any of these questions were classified as having COPD [20, 21].
Definition of BRI
The BRI was defined as follows: BRI = 364.2–365.5 × √[1 – (WC/(2π))2/(0.5 × height)2 [5]. Waist circumference and height were measured at the mobile examination centers using standardized protocols. The body measurement data were collected by trained health technicians. The health technician was assisted by a recorder during the body measurement examination. The body measurement examination rooms in each of the mobile examination centers were identical with respect to layout and equipment. Scheduled equipment calibration was performed by the health technicians and verified by supervisory staff. The Body Measures Procedures Manual includes detailed descriptions of the quality assurance and quality control measures that are used in NHANES anthropometry. These standardized protocols ensured consistency and accuracy in data collection.
Variable extraction
Demographic information for participants was extracted from the NHANES database, including socioeconomic characteristics such as gender, age, race (Mexican American, Hispanic, non-Hispanic White, non-Hispanic Black, or other), marital status (married, unmarried, divorced, or other), education level (≤ high school, > high school), and household income-to-poverty ratio (< 1.3, 1.3–3.5, or > 3.5). Additionally, lifestyle factors such as smoking status (yes or no), alcohol consumption (yes or no), hypertension, and diabetes were collected. Physical examination and laboratory test results, including height, weight, waist circumference, body mass index (BMI), total cholesterol (TC), cotinine (nicotine metabolite), white blood cells (WBC), lymphocytes (LYM), neutrophils (NE), eosinophils (EOS), red blood cells (RBC), hemoglobin (Hb), platelets (PLT), forced expiratory volume in 1 s (FEV1), and forced vital capacity (FVC), were also extracted.
Statistical methods
Given the complex sampling structure of NHANES, sample weights and stratification were applied in the analysis, in accordance with NHANES data analysis guidelines. Continuous variables are presented as means ± standard deviations, while categorical variables are presented as frequencies and 95% confidence intervals (CIs). The Chi-square test or Fisher's exact test was used for categorical variables, while t-tests and Wilcoxon rank-sum tests were employed for continuous variables. Variables with missing data exceeding 10% were excluded from the analysis. For variables with missing values between 5 and 10%, multiple imputation was applied to fill in missing values using the most appropriate dataset. Variables with less than 5% missing data were imputed with the mean value of the variable. Baseline variables showing a significance level of p < 0.05 between survivors and non-survivors were included in the multivariable models.
To minimize potential statistical bias, a weighted Cox regression model was employed for survival analysis rather than traditional regression models. The initial model was unadjusted, and Model 1 included adjustments for age, gender, race, education, marital status, household income-to-poverty ratio, smoking, alcohol consumption, and weight. Model 2 further adjusted for laboratory results (LYM, NE, EOS, RBC). Kaplan–Meier curves were employed to analyze overall survival probabilities and cardiovascular disease-specific survival probabilities across different quartiles of the BRI. Nelson-Aalen cumulative hazard function plots were employed to depict the cumulative risk of overall mortality and cardiovascular disease-specific mortality at different BRI quartile levels in the COPD population.
Receiver operating characteristic (ROC) curves were constructed to assess the specificity and sensitivity of the BRI for predicting mortality. Restricted cubic spline analysis, a widely used method, was applied to explore potential nonlinear relationships between the BRI and both all-cause and cardiovascular mortality.
Finally, subgroup analysis was conducted to examine the association between the BRI and all-cause mortality across various subgroups. After excluding participants who died within two years of the survey, the association between the BRI and all-cause mortality was reassessed. A p-value of < 0.05 was considered statistically significant. All analyses were performed using the R statistical software (version 4.3.3, http://www.R-project.org).
Results
Demographic characteristics of the COPD population
A total of 3,672 participants diagnosed with COPD were included in the study (Table 1). The average age of the COPD population was 53.17 ± 16.46 years, with 23.47% of participants aged 20–39 years, 38.25% aged 40–59 years, and 38.28% aged 60 years or older. The proportion of females was higher than that of males (males: 37.01%, females: 62.99%). Among the participants, 49.08% were married, and 44.91% belonged to middle-income households. More than half of the COPD population was non-Hispanic White (77.81%), and approximately half had attained a high school diploma or higher education (49.36%). Approximately 30% of the COPD population was overweight, and nearly 45% were obese. At the time of the interview, more than half of the participants were current alcohol users (77.5%), and approximately 70% were smokers. The average BRI was 6.01 ± 2.61. A total of 1,008 deaths were observed, 316 of which were due to cardiovascular causes. Non-survivors were more likely to be non-Hispanic White, have lower education levels, belong to middle-income households, be older, be smokers, with hypertension and diabetes, have lower body weight, and have lower lymphocyte, red blood cell, FEV1, and FVC levels, but higher neutrophil and eosinophil levels (Table 1).
Table 1.
Baseline characteristics between survivors and non-survivors
| Variable | Total (n = 3672) |
Survivor (n = 2664) |
Non-survivors (n = 1008) |
P* |
|---|---|---|---|---|
| Height (cm) | 166.81 ± 9.64 | 166.99 ± 9.54 | 166.22 ± 9.93 | 0.042 |
| Weight (kg) | 84.03 ± 23.67 | 85.05 ± 23.39 | 80.59 ± 24.29 | < 0.001 |
| Waist circumference (cm) | 102.57 ± 18.05 | 102.48 ± 17.80 | 102.88 ± 18.85 | 0.576 |
| COT (ng/mL) | 100.46 ± 143.85 | 101.28 ± 144.61 | 97.71 ± 141.20 | 0.527 |
| TC (mg/dl) | 198.83 ± 44.59 | 198.76 ± 45.68 | 199.05 ± 40.72 | 0.87 |
| WBC (K/UL) | 7.85 ± 2.48 | 7.81 ± 2.45 | 7.97 ± 2.57 | 0.103 |
| LYM (K/UL) | 2.22 ± 1.32 | 2.26 ± 1.24 | 2.10 ± 1.55 | 0.002 |
| NE (K/UL) | 4.74 ± 1.79 | 4.69 ± 1.78 | 4.93 ± 1.83 | < 0.001 |
| EOS (K/UL) | 0.23 ± 0.20 | 0.22 ± 0.17 | 0.24 ± 0.28 | 0.011 |
| RBC (K/UL) | 4.64 ± 0.48 | 4.66 ± 0.47 | 4.56 ± 0.51 | < 0.001 |
| HB (g/dL) | 14.17 ± 1.49 | 14.19 ± 1.47 | 14.09 ± 1.56 | 0.063 |
| PLT (K/UL) | 264.74 ± 70.45 | 265.37 ± 67.82 | 262.60 ± 78.61 | 0.316 |
| BRI | 6.01 ± 2.61 | 5.99 ± 2.60 | 6.10 ± 2.64 | 0.25 |
| FEV1 (ml) | 2631.70 ± 882.17 | 2736.57 ± 833.34 | 2023.68 ± 911.34 | < 0.001 |
| FVC (ml) | 3607.98 ± 1050.41 | 3697.81 ± 1004.22 | 3087.16 ± 1155.69 | < 0.001 |
| Age (years) | 53.17 ± 16.46 | 49.50 ± 15.38 | 65.51 ± 13.76 | < 0.001 |
| 20–39 (%, 95%CI) | 23.47 (21.5,25.55) | 28.53 (26.16,31.02) | 6.44 (3.97,10.29) | |
| 40–59 (%, 95%CI) | 38.25 (36.07,40.48) | 43.31 (40.74,45.93) | 21.23 (17.69,25.26) | |
| ≥ 60 (%, 95%CI) | 38.28 (36.24,40.37) | 28.16 (26.02,30.4) | 72.33 (67.8,76.45) | |
| BMI (kg/m2) | 30.11 ± 7.83 | 30.44 ± 7.80 | 29.02 ± 7.84 | < 0.001 |
| < 25 (%, 95%CI) | 27.27 (25.32,29.31) | 25.1 (22.87,27.47) | 34.57 (30.71,38.64) | |
| ≥ 25 和 < 30 (%, 95%CI) | 28.24 (26.3,30.26) | 28.33 (26.05,30.73) | 27.92 (24.5,31.61) | |
| ≥ 30 (%, 95%CI) | 44.49 (42.29,46.72) | 46.57 (43.99,49.17) | 37.52 (33.42,41.8) | |
| Gender (%, 95%CI) | < 0.001 | |||
| Male | 37.01 (34.92,39.16) | 34.75 (32.33,37.26) | 44.62 (40.54,48.77) | |
| Female | 62.99 (60.84,65.08) | 65.25 (62.74,67.67) | 55.38 (51.23,59.46) | |
| Race (%, 95%CI) | < 0.001 | |||
| Mexican | 3.1 (2.69,3.57) | 3.5 (2.99,4.09) | 1.76 (1.33,2.34) | |
| Hispanics | 4.36 (3.47,5.47) | 4.93 (3.92,6.2) | 2.44 (1.04,5.63) | |
| Non-Hispanic White | 77.81 (76.16,79.38) | 75.56 (73.58,77.43) | 85.38 (82.36,87.96) | |
| Non-Hispanic Black | 9.19 (8.39,10.05) | 9.81 (8.85,10.86) | 7.08 (5.86,8.53) | |
| Others | 5.54 (4.63,6.62) | 6.2 (5.1,7.51) | 3.33 (2.07,5.33) | |
| Education (%, 95%CI) | < 0.001 | |||
| High school and Less than high school | 50.64 (48.41,52.86) | 46.72 (44.13,49.32) | 63.82 (59.68,67.77) | |
| More than high school | 49.36 (47.14,51.59) | 53.28 (50.68,55.87) | 36.18 (32.23,40.32) | |
| Marital status (%, 95%CI) | < 0.001 | |||
| Married | 49.08 (46.86,51.31) | 49.92 (47.32,52.53) | 46.25 (42.09,50.46) | |
| Never married | 12.29 (10.9,13.82) | 14.46 (12.75,16.35) | 4.99 (3.39,7.27) | |
| Divorced | 15.17 (13.71,16.76) | 15.36 (13.65,17.25) | 14.53 (11.98,17.52) | |
| Others | 23.46 (21.66,25.36) | 20.25 (18.29,22.37) | 34.24 (30.32,38.38) | |
| Family income of poverty ratio (%, 95%CI) | < 0.001 | |||
| < 1.3 | 28.5 (26.6,30.48) | 26.51 (24.4,28.73) | 35.2 (31.21,39.41) | |
| 1.3–3.5 | 44.91 (42.71,47.12) | 44.27 (41.7,46.86) | 47.05 (42.91,51.24) | |
| > 3.5 | 26.59 (24.56,28.73) | 29.22 (26.79,31.78) | 17.75 (14.71,21.26) | |
| Alcohol use (%, 95%CI) | < 0.001 | |||
| Yes | 77.5 (75.62,79.27) | 79.39 (77.22,81.4) | 71.13 (67.29,74.69) | |
| No | 22.5 (20.73,24.38) | 20.61 (18.6,22.78) | 28.87 (25.31,32.71) | |
| Smoking status (%, 95%CI) | < 0.001 | |||
| Yes | 67.36 (65.24,69.42) | 64.93 (62.42,67.36) | 75.55 (71.6,79.1) | |
| No | 32.64 (30.58,34.76) | 35.07 (32.64,37.58) | 24.45 (20.9,28.4) | |
| Hypertension | < 0.001 | |||
| Yes | 45.41 (43.23,47.61) | 41.61 (39.12,44.15) | 58.18 (53.93,62.32) | |
| No | 54.59 (52.39,56.77) | 58.39 (55.85,60.88) | 41.82 (37.68,46.07) | |
| Diabetes | < 0.001 | |||
| Yes | 15.49 (14.06,17.03) | 13.74 (12.15,15.49) | 21.37 (18.37,24.71) | |
| No | 84.51 (82.97,85.94) | 86.26 (84.51,87.85) | 78.63 (75.29,81.63) |
COT cotinine, TC total cholesterol, WBC white blood cell, LYM lymphocyte, NE neutrophil, EOS Eosinophil, RBC Red Blood Cell, HB Hemoglobin, PLT platelet, BRI body roundness index, FEV1 Forced Expiratory Volume in 1 s, FVC Forced Vital Capacity, BMI body mass index
Association between the BRI and all-cause and cardiovascular mortality risks
Cox proportional hazards analysis revealed a significant association between the BRI and all-cause mortality risk when treated as a continuous variable. This association was observed in the unadjusted model (hazard ratio [HR] 1.06; 95% CI 1.02, 1.09), model 1 (HR 1.15; 95% CI 1.07, 1.23), and model 2 (HR 1.12; 95% CI 1.05, 1.19). Similarly, for cardiovascular mortality, Cox proportional hazards analysis showed a significant association between the BRI and cardiovascular mortality risk in the unadjusted model (HR 1.09; 95% CI 1.03, 1.15), model 1 (HR 1.19; 95% CI 1.05, 1.34), and model 2 (HR 1.14; 95% CI 1.01, 1.29) (Table 2).
Table 2.
Association Between the BRI and All-Cause and Cardiovascular Mortality (Cox Regression)
| Groups | Non-adjusted | Model 1 | Model 2 | ||||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P-Value | HR (95% CI) | P-Value | HR (95% CI) | P-Value | ||
| All-cause mortality | Continuous | 1.06 (1.02,1.09) | 0.002 | 1.15 (1.07,1.23) | < 0.001 | 1.12 (1.05, 1.19) | 0.001 |
| Q1 (≤ 4.296; n = 918) | Ref | Ref | Ref | Ref | Ref | Ref | |
| Q2 (> 4.296, ≤ 5.876; n = 918) | 0.89 (0.7,1.14) | 0.353 | 0.75 (0.59,0.95) | 0.018 | 0.75 (0.59, 0.96) | 0.02 | |
| Q3 (> 5.876, ≤ 7.68; n = 918) | 1.05 (0.82,1.33) | 0.719 | 0.97 (0.74,1.28) | 0.842 | 0.95 (0.72, 1.27) | 0.745 | |
| Q4 (> 7.68; n = 918) | 1.3 (1.01,1.68) | 0.041 | 1.33 (0.91,1.93) | 0.14 | 1.20 (0.82, 1.76) | 0.345 | |
| P for trend | 0.04 | < 0.001 | < 0.001 | ||||
| Cardiovascular mortality | Continuous | 1.09 (1.03, 1.15) | 0.002 | 1.19 (1.05, 1.34) | 0.006 | 1.14 (1.01, 1.29) | 0.035 |
| Q1 (≤ 4.296; n = 918) | Ref | Ref | Ref | Ref | Ref | Ref | |
| Q2 (> 4.296, ≤ 5.876; n = 918) | 0.92 (0.62, 1.36) | 0.676 | 0.57 (0.36, 0.9) | 0.017 | 0.62 (0.39, 0.99) | 0.047 | |
| Q3 (> 5.876, ≤ 7.68; n = 918) | 1.16 (0.8, 1.69) | 0.439 | 0.81 (0.48, 1.35) | 0.417 | 0.80 (0.47, 1.34) | 0.393 | |
| Q4 (> 7.68; n = 918) | 1.5 (0.99, 2.27) | 0.053 | 1 (0.52, 1.93) | 0.996 | 1.00 (0.52, 1.95) | 0.994 | |
| P for trend | 0.1 | < 0.001 | < 0.001 | ||||
P < 0.05 indicates statistical significance
Model 1: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, and weight
Model 2: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, weight, lymphocytes, neutrophils, eosinophils, red blood cells, forced expiratory volume in 1 s, forced vital capacity, hypertension, and diabetes
Participants were then divided into four groups based on the quartiles of the BRI: Q1 (≤ 4.296; n = 918), Q2 (4.296 < BRI ≤ 5.876; n = 918), Q3 (5.876 < BRI ≤ 7.68; n = 918), and Q4 (> 7.68; n = 918). In the analysis of all-cause mortality risk by BRI quartiles, Cox proportional hazards analysis revealed significant associations in the Q4 group in the unadjusted model, the Q1 group in model 1, and the Q1 group in model 2. For cardiovascular mortality risk by BRI quartiles, significant associations were found in the Q1 group in both model 1 and model 2 (Table 2, Fig. 2).
Fig. 2.
Cox Proportional hazards model forest plot. A Association Between the BRI and All-Cause Mortality in the Unadjusted Model, Model 1, and Model 2; B Association Between the BRI and Cardiovascular Mortality in the Unadjusted Model, Model 1, and Model 2
Nonlinear trend between the BRI and mortality in the COPD population
As shown in Fig. 3A and B, multivariable-adjusted restricted cubic spline (RCS) analysis was performed to visualize the relationship between the BRI and both all-cause and cardiovascular mortality in the COPD population. The analysis was adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, weight, hypertension, diabetes, lymphocytes, neutrophils, eosinophils, red blood cell, forced expiratory volume in 1 s, and forced vital capacity levels. A nonlinear trend was observed in the relationship between the BRI and all-cause mortality (P for nonlinear < 0.001) (Fig. 3A). In contrast, a linear trend was observed in the relationship between the BRI and cardiovascular mortality (P for nonlinear = 0.122) (Fig. 3B).
Fig. 3.
Risk curve for BRl changes using RCS. A All-Cause Mortality Risk at Different Levels of the BRI; B Cardiovascular Mortality Risk at Different Levels of the BRI
The linear trend of the BRI was assessed using restricted cubic splines to evaluate both all-cause and cardiovascular mortality in the COPD population. (Fig. 3A) All-cause mortality risk across different levels of the BRI; (Fig. 3B) Cardiovascular mortality risk across different levels of the BRI. The dashed lines represent the 95% confidence intervals (CIs). The spline analysis was adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, weight, hypertension, diabetes, lymphocytes, neutrophils, eosinophils, red blood cell, forced expiratory volume in 1 s, and forced vital capacity levels.
ROC curve analysis of the BRI
The ROC curve analysis assessing the ability of the BRI to predict mortality in COPD patients is shown in Fig. 4. The results in Fig. 4 show that the three models demonstrated better prediction performance for both all-cause mortality [0.52 (0.50–0.54) vs. 0.80 (0.78–0.81) vs. 0.81 (0.79–0.82)] and cardiovascular mortality [0.50 (0.47–0.54) vs. 0.83 (0.80–0.85) vs. 0.79 (0.76–0.81)], with the fully adjusted risk model providing the best prediction. Additionally, the cutoff value of the BRI for predicting all-cause mortality was 7.46, while the cutoff for predicting cardiovascular mortality was 5.14 (Table 3).
Fig. 4.
ROC Curves of the BRI as a Predictor of All-Cause and Cardiovascular Mortality. A BRI predicting all-cause mortality. B Adjusted risk model + BRI predicting all-cause mortality. C Fully adjusted risk model + BRI predicting all-cause mortality. D BRI predicting cardiovascular mortality. E Adjusted risk model + BRI predicting cardiovascular mortality. F Fully adjusted risk model + BRI predicting cardiovascular mortality. Adjusted risk model: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, and weight. Fully adjusted risk model: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, weight, lymphocytes, neutrophils, eosinophils, red blood cells, forced expiratory volume in 1 s, forced vital capacity, hypertension, and diabetes
Table 3.
Discrimination of the BRI in predicting all-cause and cardiovascular mortality
| Models | AUC (95% CI) | Cut-off | Sensitivity (95% CI) | Specificity (95% CI) |
|---|---|---|---|---|
| All-cause mortality | ||||
| BRI | 0.52 (0.50–0.54) | 7.46 | 0.71 (0.69—0.73) | 0.23 (0.21—0.26) |
| Cardiovascular mortality | ||||
| BRI | 0.50 (0.47–0.54) | 5.14 | 0.37 (0.36—0.39) | 0.67 (0.62—0.73) |
Kaplan–Meier curves
The Kaplan–Meier (K-M) curves show the overall survival probability and cardiovascular disease-specific survival probability for the COPD population at different BRI quartile levels. The K-M curves indicate that compared to the highest quartile group, the COPD population in the lowest BRI quartile has higher overall survival probability (P = 0.038) (Fig. 5A). Additionally, the Nelson-Aalen cumulative hazard function plots show that, over time, the cumulative risk of overall mortality and cardiovascular disease-specific mortality continuously increases (Fig. 5C, D). Compared to the lowest quartile group, the COPD population in the highest BRI quartile has higher cumulative overall mortality risk (Fig. 5C).
Fig. 5.
Kaplan–Meier Curves and Nelson-Aalen cumulative hazard function plots. A Overall Survival at Different BRI Quartiles; B Cardiovascular Disease-Specific Survival at Different BRI Quartiles; C Cumulative Overall Mortality Risk at Different BRI Quartiles; D Cumulative Cardiovascular Mortality Risk at Different BRI Quartiles
Subgroup and sensitivity analyses
Figure 6 shows a stratified analysis of overall survival probability by age, gender, education level, and household income-to-poverty ratio. The results suggest significant differences in survival curves based on gender, education level, and household income-to-poverty ratio across different BRI quartile levels in the COPD population (Fig. 6B, C, G, I).
Fig. 6.
Stratified analysis of overall survival probability at different BRI quartiles. A Overall survival probability for participants aged 20–39 years. B Overall survival probability for participants aged 40–59 years. C Overall survival probability for participants aged ≥ 60 years. D Overall survival probability for male participants. E Overall survival probability for female participants. F Overall survival probability for participants with ≤ high school education. G Overall survival probability for participants with > high school education. H Overall survival probability for participants with household income-to-poverty ratio < 1.3. I Overall survival probability for participants with household income-to-poverty ratio 1.3–3.5. J Overall survival probability for participants with household income-to-poverty ratio > 3.5
In the subgroup analysis by age, a significant difference in survival curves was observed among patients aged 40 years and older (P = 0.038, P = 0.031). The overall survival probability for the Q2 group was higher than that for the Q1 and Q4 groups (Fig. 6C).
In the analysis stratified by education level, a significant difference in survival curves was observed for patients with a high school education or higher (P = 0.041), with the Q4 group showing lower overall survival probability compared to the other groups.
In the analysis stratified by household income-to-poverty ratio, a significant difference in survival curves was observed among patients with an income-to-poverty ratio between 1.3 and 3.5 (P = 0.039), with the Q4 group again showing lower overall survival probability compared to the other groups.
Subgroup analyses revealed significant interaction effects (P < 0.05), indicating that the relationship between the BRI and all-cause mortality differs across subgroups in the COPD population. Specifically, the impact of the BRI on all-cause mortality varied by alcohol consumption status (Fig. 7).
Fig. 7.
Subgroup analysis of the BRI and all-cause mortality in the COPD population
After excluding participants who died within two years of the survey, the BRI remained associated with all-cause mortality during follow-up, with a hazard ratio (HR) of 1.10 (95% CI: 1.03, 1.18) (Table 4). Similarly, in the alcohol-consuming subgroup, the results remained robust, with an HR of 1.14 (95% CI: 1.05, 1.24) (Table 5). These subgroup and sensitivity analyses demonstrate the robustness and reliability of the models.
Table 4.
Association Between the BRI and all-cause mortality after excluding participants who died within two years of the survey
| Groups | Non-adjusted | Model 1 | Model 2 | ||||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P-Value | HR (95% CI) | P-Value | HR (95% CI) | P-Value | ||
| All-cause mortality | Continuous | 1.06 (1.02, 1.1) | 0.003 | 1.13 (1.05, 1.21) | 0.002 | 1.10 (1.03, 1.18) | 0.008 |
| Q1 | Ref | Ref | Ref | Ref | Ref | Ref | |
| Q2 | 0.86 (0.66, 1.12) | 0.259 | 0.71 (0.55, 0.92) | 0.01 | 0.71 (0.55, 0.92) | 0.011 | |
| Q3 | 1.1 (0.85, 1.43) | 0.478 | 1.02(0.75, 1.39) | 0.892 | 1.01 (0.73, 1.38) | 0.961 | |
| Q4 | 1.31 (0.99, 1.73) | 0.063 | 1.3 (0.86, 1.98) | 0.212 | 1.20 (0.78, 1.84) | 0.402 | |
| P for trend | 0.04 | < 0.001 | < 0.001 | ||||
(Cox Regression) Statistical Significance at P < 0.05 Level
Model 1: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, and weight
Model 2: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, alcohol consumption, smoking status, weight, lymphocytes, neutrophils, eosinophils, red blood cells, forced expiratory volume in 1 s, forced vital capacity, hypertension, and diabetes
Table 5.
Association between the BRI and all-cause mortality in the COPD population with alcohol consumption
| Groups | Non-adjusted | Model 1 | Model 2 | ||||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P-Value | HR (95% CI) | P-Value | HR (95% CI) | P-Value | ||
| All-cause mortality | Continuous | 1.08 (1.03, 1.12) | < 0.001 | 1.18 (1.08, 1.29) | < 0.001 | 1.14 (1.05, 1.24) | 0.003 |
| Q1 | Ref | Ref | Ref | Ref | Ref | Ref | |
| Q2 | 0.95 (0.71, 1.26) | 0.713 | 0.91 (0.73, 1.14) | 0.411 | 0.78 (0.58, 1.04) | 0.093 | |
| Q3 | 1.07 (0.8, 1.43) | 0.653 | 1.21 (0.94, 1.55) | 0.148 | 0.95 (0.66, 1.38) | 0.802 | |
| Q4 | 1.48 (1.09, 2.02) | 0.012 | 1.66 (1.18, 2.33) | 0.004 | 1.34 (0.82, 2.20) | 0.246 | |
| P for trend | 0.03 | < 0.001 | < 0.001 | ||||
(Cox Regression) Statistical Significance at P < 0.05 Level
Model 1: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, smoking status, and weight
Model 2: Adjusted for age, gender, race, education, marital status, household income-to-poverty ratio, smoking status, weight, lymphocytes, neutrophils, eosinophils, red blood cells, forced expiratory volume in 1 s, forced vital capacity, hypertension, and diabetes
Discussion
COPD is a common chronic inflammatory disease of the respiratory system, characterized by a prolonged course and recurrent acute exacerbations. Its progression significantly impacts patients'quality of life and health, leading to a considerable economic burden on both families and society. The inpatient mortality rate among elderly COPD patients is reported to range between 5 and 11% [22], while the in-hospital mortality rate for COPD patients admitted with acute respiratory failure ranges from 11% to 36.6% [23, 24]. The primary recognized risk factors for COPD include smoking and exposure to harmful gases, such as air pollution, which induce chronic airway inflammation. However, limited research exists on other factors, such as waist circumference, waist-to-hip ratio, and the BRI, and their association with COPD outcomes.
Obesity is widely recognized as a risk factor for cardiovascular diseases, diabetes, hypertension, and other serious health conditions. It is also believed to worsen the prognosis in COPD patients [25]. Studies have linked obesity to reduced chest wall compliance, decreased lung volumes, impaired airway function, diaphragm dysfunction, and arterial hypoxemia [26]. Specifically, obesity can decrease expiratory reserve volume (ERV) and functional residual capacity (FRC), lower flow rates at low volumes, and increase residual volume (RV) as a proportion of total lung capacity (TLC) [27]. Further studies suggest that reduced chest wall compliance is caused by the accumulation of fat around the chest and abdomen in obesity [28]. Therefore, abdominal fat plays a key role in the decline of lung function in COPD patients. However, relying solely on body weight or BMI to assess obesity may not be accurate, as these measures do not account for the distribution of fat and non-fat tissues [29].
Body measurements, such as waist circumference, can reflect the degree of obesity and have predictive value for the incidence and severity of cardiovascular and other chronic diseases [30]. Waist circumference, particularly abdominal or visceral fat, is a key feature of metabolic syndrome and is strongly correlated with various chronic diseases. Therefore, waist circumference is essential in disease prognosis assessments [31]. Some studies suggest that waist circumference affects lung ventilation mechanics by limiting diaphragm movement [32]. Furthermore, waist circumference has shown better predictive value for the FEV1/FVC ratio [33, 34]. Research by Ma et al. [35] shows that abdominal obesity is closely related to lower levels of FEV1 and FVC, and an increased risk of COPD development. C. Maiolo et al. [28] demonstrated that fat tissue is negatively correlated with lung function, whereas non-fat tissue is positively correlated, and patients with a waist-to-hip ratio (W/H) ≥ 0.95 have significantly lower FVC and FEV1.Adipose tissue is not only an important source of a variety of signaling molecules, adiponectin, but also a source of stem cells that can in tissue repair, two hormones produced by adipose tissue, leptin and adiponectin [36]. Leptin has pro-inflammatory properties [37]. Elevated leptin levels may modulate immune responses in the airways by triggering robust pro-inflammatory responses or skewing cellular responses toward a type 1 helper phen [38, 39]. Leptin may also increase bronchial hyperreactivity [40].
This study examined the relationship between the BRI and both all-cause and cardiovascular mortality in the COPD population. A total of 3,672 COPD participants were included, consisting of 1,517 males and 2,155 females. The average age of the study population was 53.17 ± 16.46 years. Baseline characteristics showed that non-survivors were typically older, non-Hispanic White, had lower education levels, lower household income, were more likely to smoke, and had lower levels of lymphocytes and red blood cells, but higher levels of neutrophils and eosinophils. Multivariable Cox regression analysis revealed that the BRI was associated with both all-cause and cardiovascular mortality in COPD patients.
The Kaplan–Meier (K-M) curves and Nelson-Aalen cumulative risk function analysis demonstrated that the overall survival probability in the Q4 group (highest BRI quartile) was significantly lower, and the cumulative incidence of all-cause mortality increased with higher BRI quartiles. ROC curve analysis indicated that the fully adjusted risk model outperformed others in predicting both all-cause and cardiovascular mortality. Additionally, subgroup analysis found that the effect of BRI on all-cause mortality in COPD population was different between the subgroups of whether drink (P < 0.05). However, the sensitivity analysis suggested that the results were still robust in the study population who drank after excluding the participants who within 2 years, indicating the stability of the results. Most scholars in the past believed that drinking could lead to weight gain and trigger abdominal obesity, BRI increased [41]. This is related to the fact that drinking increases fasting TG and VLDL levels, because ethanol can stimulate the release of fatty acids from adipose tissue promote the synthesis of TG and VLDL in the liver, and slow down the clearance rate of VLDL and chylomicrons from the blood [42, 43]. These results align with the findings of most previous studies. For example, Li et al. [44] found that high waist circumference is a risk factor for COPD, while other studies suggest that assessing body fat distribution may aid in the clinical evaluation of COPD mortality risk [45].Moreover, previous studies have found that metabolic syndrome [46], hyperuricemia [47] and left ventricular hypertrophy [48] have been proved to be closely related to BRI, indicating that BRI has a wide clinical significance.
Our study further confirmed that the BRI is associated with both all-cause and cardiovascular mortality in COPD patients. Subgroup and sensitivity analyses demonstrated the robustness and reliability of the model. At present, most studies on the effect of obesity on COPD use BMI and waist circumference, but BRI is more accurate [49, 50].The BRI offers distinct advantages in analyzing COPD mortality. This finding provides a new perspective on evaluating mortality risk in COPD patients. However, this study has some limitations:(1)The classification of all-cause mortality and cardiovascular disease (CVD) mortality in NHANES is based entirely on ICD codes. While this method has been validated by the U.S. Centers for Disease Control and Prevention (CDC) and is widely used in CDC reports, there remains a possibility of misclassification of cause of death.we plan to combine more objective lung function data or explore other data sources that can further validate self-reported diagnoses. (2)The present study was a cross-sectional study, and cross-sectional study only reveals the correlation and cannot make causal inference. In addition, selection bias other effects may exist.In the future, it is necessary to carry out high-quality multi-center, large sample randomized controlled trials to explore the relationship and mechanism between BRI and COPD mortality.
Conclusion
The BRI is associated with both all-cause and cardiovascular mortality in the COPD population, with the cumulative incidence of all-cause mortality increasing with higher BRI quartiles. Further research is needed to investigate interventions aimed at improving the BRI to reduce mortality in the COPD population.
Acknowledgements
We would also like to thank everyone who has helped with our research.
Authors’ contributions
RKC, YLC, QHC, ZJC are the guarantor of the manuscript and take responsibility for the content of this manuscript. LJZ, HMC and BZZ contributed to the design of the study. QHC, YHS, HL, YTY and XL were involved in the data analysis. XYL, JFC and HMC contributed to the acquisition of primary data. RKC and YLC wrote the initial draft of the manuscript. LJZ, BZZ and HMC contributed significantly to the revision of the manuscript. All authors read and approved the final manuscript.
Funding
This study was funded by the Natural Science Foundation of Basic and Applied Basic Research Fund of Guangdong Province (2022A1515012375), the Guangdong Medical Research Fund Project (A2024728, A2024723), the Science and Technology Development Special Project of Zhanjiang City (2022A01142, 2022A01110, 2021A05086, 2021A05088).
Data availability
The data used to support the findings of this study are available from the corresponding author upon request.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publications
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Riken Chen, Yuli Cai, Qinghua Chen and Zhaojun Chen contributed equally to this work.
Contributor Information
Lijuan Zeng, Email: ZLJ13729063609@163.com.
Huimin Chen, Email: gzychm@163.com.
Baozhi Zhang, Email: 87992786@qq.com.
References
- 1.Byers BW, et al. The effect of pulmonary rehabilitation on carotid chemoreceptor activity and sensitivity in chronic obstructive pulmonary disease. J Appl Physiol (1985). 2019;127(5):1278–87. [DOI] [PubMed] [Google Scholar]
- 2.Barnes PJ. Targeting cytokines to treat asthma and chronic obstructive pulmonary disease. Nat Rev Immunol. 2018;18(7):454–66. [DOI] [PubMed] [Google Scholar]
- 3.Hanson C, LeVan T. Obesity and chronic obstructive pulmonary disease: recent knowledge and future directions. Curr Opin Pulm Med. 2017;23(2):149–53. [DOI] [PubMed] [Google Scholar]
- 4.Lozano R, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Thomas DM, et al. Relationships between body roundness with body fat and visceral adipose tissue emerging from a new geometrical model. Obesity. 2013;21(11):2264–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kang SM, et al. Android fat depot is more closely associated with metabolic syndrome than abdominal visceral fat in elderly people. PLoS One. 2011;6(11):e27694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cherniack RM, Raber MB. Normal standards for ventilatory function using an automated wedge spirometer. Am Rev Respir Dis. 1972;106(1):38–46. [DOI] [PubMed] [Google Scholar]
- 8.Cotes JE, et al. Average normal values for the forced expiratory volume in white Caucasian males. Br Med J. 1966;1(5494):1016–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Knudson RJ, et al. The maximal expiratory flow-volume curve. Normal standards, variability, and effects of age. Am Rev Respir Dis. 1976;113(5):587–600. [DOI] [PubMed] [Google Scholar]
- 10.Vague J. Sexual differentiation; factor determining forms of obesity. Presse Med (1893). 1947;55(30):339. [PubMed] [Google Scholar]
- 11.Vgontzas AN. Sleep apnea and daytime sleepiness and fatigue: relation to visceral obesity, insulin resistance, and hypercytokinemia. J Clin Endocrinol Metab. 2000;85(3):1151–8. [DOI] [PubMed] [Google Scholar]
- 12.Dhaliwal SS, Welborn TA. Measurement error and ethnic comparisons of measures of abdominal obesity. Prev Med. 2009;49(2–3):148–52. [DOI] [PubMed] [Google Scholar]
- 13.Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature. 2006;444(7121):840–6. [DOI] [PubMed] [Google Scholar]
- 14.Li G, et al. The feasibility of two anthropometric indices to identify metabolic syndrome, insulin resistance and inflammatory factors in obese and overweight adults. Nutrition. 2019;57:194–201. [DOI] [PubMed] [Google Scholar]
- 15.Zhao W, et al. Relationship between body roundness index and risk of type 2 diabetes in japanese men and women: a reanalysis of a cohort study. Int J Endocrinol. 2021;2021:4535983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen Y, et al. Waist circumference is associated with pulmonary function in normal-weight, overweight, and obese subjects. Am J Clin Nutr. 2007;85(1):35–9. [DOI] [PubMed] [Google Scholar]
- 17.Lee SJ, Arslanian SA. Cardiorespiratory fitness and abdominal adiposity in youth. Eur J Clin Nutr. 2007;61(4):561–5. [DOI] [PubMed] [Google Scholar]
- 18.Johnson CL, et al. National health and nutrition examination survey: sample design, 2011–2014. Vital Health Stat 2. 2014;162:1–33. [PubMed] [Google Scholar]
- 19.Qiu S, Jiang Q, Li Y. The association between pan-immune-inflammation value and chronic obstructive pulmonary disease: data from NHANES 1999–2018. Front Physiol. 2024;7(15):1440264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Huang Q, et al. Association between ethylene oxide exposure and prevalence of COPD: evidence from NHANES 2013–2016. Sci Total Environ. 2023;885:163871. [DOI] [PubMed] [Google Scholar]
- 21.Mai S, et al. Controlling nutritional status score in the prediction of cardiovascular disease prevalence, all-cause and cardiovascular mortality in chronic obstructive pulmonary disease population: NHANES 1999–2018. BMC Pulm Med. 2024;24(1):356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gunen H, et al. Factors affecting survival of hospitalised patients with COPD. Eur Respir J. 2005;26(2):234–41. [DOI] [PubMed] [Google Scholar]
- 23.Seneff MG, et al. Hospital and 1-year survival of patients admitted to intensive care units with acute exacerbation of chronic obstructive pulmonary disease. JAMA. 1995;274(23):1852–7. [PubMed] [Google Scholar]
- 24.Connors AJ, et al. Outcomes following acute exacerbation of severe chronic obstructive lung disease. The SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments). Am J Respir Crit Care Med. 1996;154(4 Pt 1):959–67. [DOI] [PubMed] [Google Scholar]
- 25.DeLapp DA, et al. Patients with obesity have better long-term outcomes after hospitalization for COPD exacerbation. COPD. 2020;17(4):373–7. [DOI] [PubMed] [Google Scholar]
- 26.Ray CS, et al. Effects of obesity on respiratory function. Am Rev Respir Dis. 1983;128(3):501–6. [DOI] [PubMed] [Google Scholar]
- 27.Cerveri I, et al. Reference values of arterial oxygen tension in the middle-aged and elderly. Am J Respir Crit Care Med. 1995;152(3):934–41. [DOI] [PubMed] [Google Scholar]
- 28.Maiolo C, Mohamed EI, Carbonelli MG. Body composition and respiratory function. Acta Diabetol. 2003;40(Suppl 1):S32–8. [DOI] [PubMed] [Google Scholar]
- 29.Balasubramanian P, et al. Obesity-induced cognitive impairment in older adults: a microvascular perspective. Am J Physiol Heart Circ Physiol. 2021;320(2):H740–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Karam JG, et al. Medical management of obesity: an update. Minerva Endocrinol. 2007;32(3):185–207. [PubMed] [Google Scholar]
- 31.Gasparovic K, et al. Diagnostic accuracy, sensitivity, and specificity of CT pulmonary artery to aorta diameter ratio in screening for pulmonary hypertension in end-stage COPD patients. Croat Med J. 2021;62(5):446–445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Barnes PJ, Celli BR. Systemic manifestations and comorbidities of COPD. Eur Respir J. 2009;33(5):1165–85. [DOI] [PubMed] [Google Scholar]
- 33.Ghobadi H, et al. The Relationship between COPD Assessment Test (CAT) Scores and Severity of Airflow Obstruction in Stable COPD Patients. Tanaffos. 2012;11(2):22–6. [PMC free article] [PubMed] [Google Scholar]
- 34.Gibson GJ. Obesity, respiratory function and breathlessness. Thorax. 2000;55 Suppl 1(Suppl 1):S41-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ma YL, Zhao HJ, Su YH. Association between waist circumference change and incident chronic obstructive pulmonary disease among Chinese adults: a 10-year cohort study. Sci Rep. 2022;12(1):18402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Guerre-Millo M. Adipose tissue hormones. J Endocrinol Invest. 2002;25(10):855–61. [DOI] [PubMed] [Google Scholar]
- 37.Claycombe K, King LE, Fraker PJ. A role for leptin in sustaining lymphopoiesis and myelopoiesis. Proc Natl Acad Sci U S A. 2008;105(6):2017–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Mancuso P, et al. Leptin augments alveolar macrophage leukotriene synthesis by increasing phospholipase activity and enhancing group IVC iPLA2 (cPLA2gamma) protein expression. Am J Physiol Lung Cell Mol Physiol. 2004;287(3):L497-502. [DOI] [PubMed] [Google Scholar]
- 39.Dixon AE, et al. Relationship of adipokines with immune response and lung function in obese asthmatic and non-asthmatic women. J Asthma. 2011;48(8):811–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Sideleva O, et al. Obesity and asthma: an inflammatory disease of adipose tissue not the airway. Am J Respir Crit Care Med. 2012;186(7):598–605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Wannamethee SG, Shaper AG. Alcohol, body weight, and weight gain in middle aged men. Am I Clin Nutr. 2003;77:1312–7. [DOI] [PubMed] [Google Scholar]
- 42.Wakabayashi I, Kobaba-Wakabayashi R, Masuda H. Relation of drinking alcohol to atherosclemtic risk in type 2 diabetes. Diabetes Care. 2002;25:1223–8. [DOI] [PubMed] [Google Scholar]
- 43.Arima H, Kiyohara Y, Kato I, et al. Alcohol reduces insulin—hypertension relationship in a general population: the hisayama study. J Clin Epidemiol. 2002;55:863–9. [DOI] [PubMed] [Google Scholar]
- 44.Li JR, et al. Clinical value of TS90% in evaluation of hypoxemia in patients with obstructive sleep apnea/hypoventilation syndrome. Zhonghua Yi Xue Za Zhi. 2005;85(44):3115–7. [PubMed] [Google Scholar]
- 45.Marquis K, et al. Midthigh muscle cross-sectional area is a better predictor of mortality than body mass index in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2002;166(6):809–13. [DOI] [PubMed] [Google Scholar]
- 46.Hertroijs DFL, et al. Relevant patient characteristics for guiding tailored integrated diabetes primary care: a systematic review. Prim Health Care Res Dev. 2018;19(5):424–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Funakoshi S, et al. Analysis of factors influencing pancreatic beta-cell function in Japanese patients with type 2 diabetes: association with body mass index and duration of diabetic exposure. Diabetes Res Clin Pract. 2008;82(3):353–8. [DOI] [PubMed] [Google Scholar]
- 48.Unger RH, et al. Lipid homeostasis, lipotoxicity and the metabolic syndrome. Biochim Biophys Acta. 2010;1801(3):209–14. [DOI] [PubMed] [Google Scholar]
- 49.Cheng H, et al. Worldwide epidemiology of diabetes-related end-stage renal disease, 2000–2015. Diabetes Care. 2021;44(1):89–97. [DOI] [PubMed] [Google Scholar]
- 50.Klein S, et al. Why does obesity cause diabetes? Cell Metab. 2022;34(1):11–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data used to support the findings of this study are available from the corresponding author upon request.