Association between weight-adjusted waist index and chronic obstructive pulmonary disease

TL;DR

This study found that a new obesity measure called weight-adjusted waist index (WWI) is linked to a higher risk of COPD in U.S. adults.

Contribution

The study introduces WWI as a novel obesity metric with better predictive power for COPD than traditional measures.

Findings

WWI was significantly associated with COPD prevalence (OR = 1.30).

WWI showed a dose-response relationship with COPD across quartiles.

WWI had better predictive performance (AUC = 0.662) than conventional obesity indices.

Abstract

This study aimed to investigate the association between the weight-adjusted waist index (WWI), a novel obesity metric, and the prevalence of chronic obstructive pulmonary disease (COPD) in a nationally representative sample of U.S. adults, and to compare its predictive utility for COPD against conventional obesity indices.

This cross-sectional study utilized data from the 2017–2020 National Health and Nutrition Examination Survey (NHANES). COPD diagnosis was based on self-report. The association between WWI and COPD was investigated using multivariable logistic regression models, adjusting for key covariates including age, gender, race/ethnicity, smoking status, hypertension, and diabetes. Restricted cubic splines (RCS) were used to explore potential non-linear relationships. Receiver operating characteristic (ROC) curves were used to assess WWI’s predictive performance. All statistical analyses were conducted using R software, accounting for the complex survey design and weighting.

This study comprised 3,111 participants, among whom the prevalence of COPD was 8.5%. The findings indicated a significant positive association between WWI and the prevalence of COPD (OR = 1.30, 95% CI: 1.02–1.66). When analyzed by quartiles, a significant positive dose-response relationship was observed (P for trend = 0.031). Furthermore, receiver operating characteristic (ROC) analysis revealed that WWI had significantly better predictive performance for COPD (Area Under the Curve [AUC] = 0.662) than conventional obesity indices.

Our findings suggest a significant positive association between WWI and the self-reported prevalence of COPD. WWI shows promise as a simple, non-invasive anthropometric tool that may aid in identifying individuals with higher odds of having COPD in clinical and public health settings.

Figures

Flowchart of participant selection from the National Health and Nutrition Examination Survey (NHANES) 2017–2020.Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease. Total excluded represents the number of unique participants with missing data in at least one variable. The sum of missing values for each variable is greater than the total excluded due to overlapping missingness.

Dose-response relationship between WWI and the odds of COPD.The solid blue line represents the odds ratio with WWI as a continuous variable, and the shaded area indicates the 95% confidence intervals. The histogram displays the population distribution of WWI values. The analysis was adjusted for age, gender, race/ethnicity, education level, marital status, smoking status, physical activity, alcohol consumption, hypertension, and diabetes.

Subgroup analysis demonstrating the association between WWI and the odds of COPD across various demographic and clinical characteristics.All analyses were based on the fully adjusted Model III. The consistent direction of the association across most subgroups supports the robustness of the main findings. OR, odds ratio; CI, confidence interval.

ROC curves comparing the predictive performance of different obesity indices for COPD.Abbreviations: AUC, Area Under the Curve; BMI, body mass index; COPD, chronic obstructive pulmonary disease; ROC, Receiver Operating Characteristic; WC, waist circumference; WWI, weight-adjusted waist index.

Full text

Introduction

Chronic obstructive pulmonary disease (COPD) is a highly prevalent condition that is both preventable and treatable [1]. It is primarily characterized by persistent airflow obstruction and respiratory symptoms, which commonly result from prolonged exposure to harmful particles or gases [2]. Prolonged exposure leads to significant structural changes in the airways and lung parenchyma [3]. Common symptoms of COPD include breathlessness, a persistent cough, and increased mucus production [4]. Furthermore, many patients with COPD have comorbidities that increase disability and mortality [5]. COPD represents a major global health challenge and, according to the latest Global Burden of Disease (GBD) 2021 study, it remains one of the leading causes of death worldwide [6].

Among the various modifiable risk factors for COPD, obesity has emerged as a significant contributor to its pathophysiology and burden [7]. However, standard anthropometric measures like Body Mass Index (BMI) have notable limitations. BMI lacks the ability to separate fat from muscle mass, does not effectively capture the distribution of adipose tissue, and its applicability may vary due to differences in race and gender [8–12]. To overcome these limitations, the weight-adjusted waist index (WWI) has been introduced as a more accurate metric for assessing central obesity [13,14]. The WWI is derived by taking the ratio of waist circumference to the square root of body weight, offering a more precise estimation of abdominal fat accumulation [15,16].

The utility of WWI is supported by evidence demonstrating its strong association with increased abdominal fat and its greater stability across diverse populations [17,18]. Furthermore, WWI has exhibited more robust connections with metabolic abnormalities, such as insulin resistance and proteinuria, in comparison to BMI, suggesting it may act as a more effective marker for assessing the risk of obesity-related metabolic disorders [19,20]. These obesity-driven pathologies, particularly the chronic low-grade inflammation and mechanical lung compression associated with central adiposity, offer a direct mechanistic pathway to the development and progression of COPD [21]. It is important to note that, unlike BMI, WWI does not currently have universally established cut-off values for diagnosing obesity. While some studies have proposed population- and outcome-specific thresholds [22], no universally accepted cut-offs exist. Given these specificities and the multi-ethnic nature of the National Health and Nutrition Examination Survey (NHANES) population, our study analyzes WWI as a continuous variable to evaluate its associated health risks.

While the link between general obesity and COPD is known, the specific association between the WWI and COPD, particularly within a large and diverse national population, has not been thoroughly investigated. Therefore, this study aims to address this gap by investigating the association between WWI and the prevalence of COPD, characterizing its dose-response relationship, and comparing its predictive utility against conventional anthropometric indices, using data from the 2017–2020 cycles of NHANES, encompassing a final sample of 3,111 adult participants.

Methods

Study design and sample

This study utilized data from the 2017–2020 NHANES, accessed by the authors starting in December 2024, to conduct a cross-sectional investigation aimed at exploring the potential association between the WWI and COPD. NHANES is a comprehensive survey conducted across the United States to evaluate the health and nutritional status of individuals living outside of institutional settings. It provides comprehensive data on health conditions, dietary intake, and demographic characteristics. NHANES employs an advanced, stratified, and multistage sampling approach, ensuring that the chosen participants accurately represent the broader population of the United States, thus enabling the creation of nationally representative health statistics. Comprehensive details on NHANES study methodology, data, and codebooks are publicly accessible on the official website: https://www.cdc.gov/nchs/nhanes/. This analysis used the publicly available, de-identified data files, and the authors did not have access to participants’ personally identifiable information.

The study protocol for NHANES was approved by the National Center for Health Statistics (NCHS) Research Ethics Review Board, and all participants provided written informed consent prior to data collection. This study utilized these publicly available, de-identified data for secondary analysis. The initial sample included 15,560 individuals who contributed valid data to the NHANES survey conducted between 2017 and 2020. To maintain data accuracy and uphold the rigor of the study, individuals with missing key health indicators (such as waist circumference, body weight, and COPD diagnosis data) or incomplete covariate information were excluded. The specific process for selecting the sample and the criteria for exclusion are comprehensively outlined in Fig 1. After applying these exclusion criteria, a final analytical sample of 3,111 participants was established. This study employed a complete case analysis, meaning only participants with complete data for all included variables were retained.

Flowchart of participant selection from the National Health and Nutrition Examination Survey (NHANES) 2017–2020.Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease. Total excluded represents the number of unique participants with missing data in at least one variable. The sum of missing values for each variable is greater than the total excluded due to overlapping missingness.

Additionally, data processing and analysis in this study adhered to the NHANES statistical guidelines, incorporating weighting measures to address biases introduced by the sampling design. These measures helped to ensure the inferential validity and statistical reliability of the results, enabling the study’s conclusions to more precisely portray the overall health status and trends of the U.S. population.

Exposure variable: WWI

Within this analysis, WWI was the main exposure variable, determined by calculating the ratio of waist circumference (cm) to the square root of body weight (kg). Data on waist circumference and body weight were sourced from the “Body Measurements” section within the NHANES examination dataset. To ensure measurement accuracy, all measurements were conducted by professionally trained healthcare personnel following standardized procedures. During weight measurement, participants wore examination clothing, stood barefoot, and maintained a relaxed posture to ensure consistency and precision. Following the official NHANES protocol, waist circumference was measured at the uppermost lateral border of the right iliac crest, with the tape positioned horizontally around the abdomen. Body weight was measured to the nearest 0.1 kg using a calibrated digital scale.

To thoroughly examine its link with obesity levels and related health risks, WWI was first treated as a continuous variable and then divided into four quartiles (Q1-Q4).

Outcome variable: COPD

The outcome variable in this study was the presence of COPD, which was determined through self-report. Data were sourced from the NHANES “Medical Conditions” questionnaire file. Participants were classified as having COPD if they answered “yes” to the question: “Ever told you had COPD, emphysema, Chronic Bronchitis (ChB)?”. While post-bronchodilator spirometry is the gold standard for clinical diagnosis, self-reported diagnoses are a common and accepted method for assessing disease prevalence in large-scale epidemiological surveys like NHANES due to their feasibility and have demonstrated reasonable validity [23]. It is important to acknowledge, however, that this approach is inherently prone to misclassification, as participants may confuse conditions like acute and chronic bronchitis, and is also subject to recall bias [24,25]. Such non-differential misclassification would likely attenuate the true association, biasing our results toward the null [26].

Covariates

Based on a review of previous literature and their clinical relevance, we selected a range of potential confounders [27–30]. Demographic data included age (continuous), gender (male or female), and race/ethnicity (categorized as Mexican American, Non-Hispanic White, Non-Hispanic Black, or Other Race). Socioeconomic and lifestyle factors included education level (grouped as Below High School or High School and Above), marital status (Married/Living with partner or Single), smoking status (current smoker or not), alcohol consumption (current drinker or not), and physical activity. Physical activity was dichotomized based on participants’ self-report. Individuals were classified as ‘active’ if they reported engaging in at least 10 continuous minutes of either moderate- or vigorous-intensity recreational activities in a typical week. Those who reported engaging in neither were classified as ‘inactive’.

Key comorbidities were also ascertained. Hypertension was defined as a self-reported physician diagnosis, an average systolic blood pressure ≥140 mmHg, or an average diastolic blood pressure ≥90 mmHg. Diabetes was defined by a self-reported physician diagnosis or a fasting plasma glucose level ≥7.0 mmol/L. The classification and coding of all categorical variables were performed in strict accordance with the official codebooks and guidelines provided in the NHANES documentation for the respective survey years.

Statistical analyses

This study employed a detailed statistical approach to explore the relationship between WWI, considered as a continuous variable, and COPD, categorized as a binary variable. Participant characteristics were summarized with continuous variables presented as mean ± standard deviation (SD) and categorical variables as proportions. Participants were divided into two distinct groups based on their COPD diagnosis. To assess differences in continuous data between these groups, weighted t-tests were performed. For categorical data, weighted chi-square tests were used to evaluate the variations.

The association between WWI and COPD was assessed using three multivariable logistic regression models. Model I was the crude, unadjusted model. Model II was adjusted for core demographic variables: age, gender, and race/ethnicity. Model III, our fully adjusted model, was adjusted for several covariates. These variables were selected based on their established roles as major risk factors for COPD and important potential confounders as supported by existing literature, and included education level, marital status, smoking status, alcohol consumption, physical activity, hypertension, and diabetes. This hierarchical approach was chosen to systematically evaluate the association while incrementally controlling for different layers of confounding.

To further explore the dose-response relationship, restricted cubic splines (RCS) were fitted to the multivariable logistic regression model. This method allowed for a flexible examination of the potential non-linear association between continuous WWI and the odds of COPD, and the linearity assumption was assessed by testing the significance of the non-linear spline terms. To assess the possible moderating influences of various factors, subgroup analyses were performed, stratified by age, gender, race, education level, marital status, smoking status, physical activity, hypertension, and diabetes. The statistical significance of any effect modification was formally tested by incorporating an interaction term between WWI and each subgroup variable into the model. All statistical analyses were conducted using R software (version 4.4.1), taking into account the complex sampling design and weighting applied in NHANES data. The ‘survey’, ‘rms’, and ‘pROC’ packages were used for these analyses. Statistical significance was determined at a threshold of *P *< 0.05.

Results

Baseline characteristics of participants

The baseline characteristics of the 3,111 participants included in the final analysis are presented in Table 1. The overall prevalence of COPD was 8.5% (N = 263). Compared with participants without COPD, those with the condition were significantly older. The proportion of Non-Hispanic White individuals was higher in the COPD group, which was also characterized by a greater prevalence of current smokers and physical inactivity. Similarly, the prevalence of both hypertension and diabetes was significantly higher among participants with COPD (P < 0.001 for both). Furthermore, participants with COPD had significantly higher mean values for WWI, BMI, and WC.

Table 1: Baseline Characteristics of Participants in the NHANES 2017–2020, Categorized by COPD Status.

Association between WWI and COPD

After full adjustment in Model III, WWI remained significantly associated with COPD (Table 2). When analyzed as a continuous variable, each unit increase in WWI was associated with a 30% increase in the odds of having COPD (OR = 1.30, 95% CI: 1.02–1.66, P = 0.034). In the quartile analysis, participants in the highest WWI quartile (Q4) had an 86% increase in the odds of COPD compared to those in the lowest quartile (Q1) (OR = 1.86, 95% CI: 1.11–3.11, P = 0.022), and the test for trend across quartiles was significant (P for trend = 0.031).

Table 2: Multivariable logistic regression models for the association between WWI and COPD.

Dose-response relationship analysis

To further explore the dose-response relationship between WWI and COPD, we performed RCS analysis based on the fully adjusted Model III. The RCS analysis revealed a significant overall association between WWI and the odds of having COPD (P for overall < 0.001), while the test for non-linearity was not statistically significant (P for non-linearity = 0.218), indicating a predominantly linear relationship between WWI and the odds of COPD across the observed range of WWI values (Fig 2). The curve demonstrates a consistent positive association, with the odds ratio for COPD steadily increasing as WWI values rise.

Dose-response relationship between WWI and the odds of COPD.The solid blue line represents the odds ratio with WWI as a continuous variable, and the shaded area indicates the 95% confidence intervals. The histogram displays the population distribution of WWI values. The analysis was adjusted for age, gender, race/ethnicity, education level, marital status, smoking status, physical activity, alcohol consumption, hypertension, and diabetes.

Subgroup analysis

To assess the robustness of our findings, we conducted subgroup analyses across various demographic and clinical characteristics, which also served as sensitivity analyses (Fig 3). The positive direction of the association between WWI and the odds of COPD was largely consistent across the majority of subgroups, supporting the stability of our main findings. Although statistical significance was not reached in some smaller subgroups, likely due to reduced statistical power, the overall consistency reinforces that the relationship between WWI and COPD is robust across diverse population characteristics.

Subgroup analysis demonstrating the association between WWI and the odds of COPD across various demographic and clinical characteristics.All analyses were based on the fully adjusted Model III. The consistent direction of the association across most subgroups supports the robustness of the main findings. OR, odds ratio; CI, confidence interval.

Interaction testing revealed that only hypertension status significantly modified this association (P for interaction = 0.007). Specifically, the association was stronger and statistically significant among participants without hypertension (OR = 1.81, 95% CI: 1.31–2.50) compared to those with hypertension (OR = 1.23, 95% CI: 0.94–1.62). No significant interactions were detected for the other variables (all P for interaction > 0.05), further supporting the general consistency of the WWI-COPD association.

Model performance evaluation

To compare the predictive value of WWI with traditional anthropometric indices for COPD, we performed ROC curve analyses (Fig 4). As detailed in Table 3, all four obesity indicators demonstrated some predictive ability, with WWI yielding the largest Area Under the Curve (AUC) of 0.662 (95% CI: 0.628–0.697). To formally test for statistical differences, we used the DeLong test to compare the AUC of WWI against the other indices. The results showed that WWI had a higher AUC compared to the other indices, and the DeLong test confirmed these differences were statistically significant for BMI (P < 0.001), WC (P = 0.035), and Weight (P < 0.001). These findings suggest that WWI has a superior predictive performance for identifying COPD compared to other commonly used obesity measures in this population.

Table 3: Predictive performance of different obesity indices for COPD.

ROC curves comparing the predictive performance of different obesity indices for COPD.Abbreviations: AUC, Area Under the Curve; BMI, body mass index; COPD, chronic obstructive pulmonary disease; ROC, Receiver Operating Characteristic; WC, waist circumference; WWI, weight-adjusted waist index.

Furthermore, we calculated the Integrated Discrimination Improvement (IDI) and Net Reclassification Improvement (NRI) to assess the incremental value of the models. Both metrics confirmed a significant improvement in performance for the adjusted models compared to the unadjusted model (all P < 0.001; Table 4).

Table 4: Model performance evaluation of different models for COPD.

Discussion

In this study of 3,111 participants, a cross-sectional analysis revealed a statistically significant positive association between WWI and COPD. Further examination through stratified subgroup analyses indicated that this positive association was generally robust across most key demographic and clinical subgroups, suggesting a broad applicability of our main finding. This general consistency underscores the potential role of WWI as an independent and stable indicator for assessing the odds of having COPD, thereby highlighting the significant contribution of central obesity in the pathogenesis of the disease. These findings suggest that managing central obesity, as measured by WWI, may be an important component of public health strategies aimed at mitigating the burden of COPD.

A key finding of our study is that WWI is not only associated with COPD but also appears to be a superior predictor compared to traditional anthropometric measures. Two key findings from our analysis support this conclusion. First, the association between WWI and COPD remained robust even after full adjustment for a wide range of confounders, with each unit increase in WWI corresponding to a 30% increase in the odds of having COPD (OR = 1.30; 95% CI: 1.02–1.66). Second, our formal comparison of predictive models using ROC analysis demonstrated that WWI had a statistically superior predictive performance for identifying COPD compared to other common indices, with the most pronounced improvement seen over BMI and Weight (P < 0.001 for both). While traditional indices like BMI are valuable, they cannot distinguish between fat and muscle mass. WWI, by integrating both waist circumference and body weight, appears to better capture the specific obesity phenotype linked to COPD [31–33]. This enhanced, statistically validated performance suggests that WWI warrants further investigation as a potentially more informative clinical and epidemiological tool [15].

Further exploration of our findings revealed several nuances. The analysis of the dose-response relationship using restricted cubic splines confirmed a significant, positive, and broadly linear association between WWI and the odds of COPD (P for overall < 0.001; P for non-linear = 0.218). This linear trend suggests a consistent increase in the odds of having COPD with each increment in WWI. Additionally, the significant interaction we observed with hypertension (P for interaction = 0.007) suggests that the impact of central obesity on COPD prevalence may be modified by the presence of this comorbidity. This finding warrants further investigation to understand how these factors synergize and to potentially tailor preventive strategies for high-risk subgroups. As this was an exploratory finding, we can only hypothesize about the underlying mechanism. One possible explanation is that in patients with established hypertension, the underlying vascular and systemic inflammatory pathways are already highly active, potentially masking the additional, relative contribution of central obesity to COPD risk [34].

Previous research has extensively investigated the risk factors associated with COPD. For instance, Holtjer et al.‘s umbrella review, which encompassed 75 reviews, summarized 45 risk factors for COPD, including a high BMI [35]. Beijers et al. emphasized the critical role of dietary and nutritional strategies in both the prevention and management of COPD [36]. Furthermore, Brock et al. elucidated the notable mechanical effects of obesity on lung function and its potential link to pulmonary diseases, thus offering a plausible explanation for the association between obesity and lung diseases [37]. Another sizable prospective study conducted among the general Japanese population also demonstrated a significant correlation between BMI, changes in body weight, and mortality from COPD [38].

Research indicates that there is a significant association between obesity, as determined by BMI or WWI, and the severity of COPD, as well as its overall prognosis. Obesity is closely interconnected with metabolic syndrome, a condition that can intensify the pathophysiological mechanisms of COPD, particularly through inflammatory responses and oxidative stress [39–42]. The systemic inflammation found in individuals with obesity can further contribute to the decline in lung function, thus exacerbating airway obstruction [43].

The association between central obesity, as captured by WWI, and COPD is biologically plausible and likely multifactorial. A key putative mechanism is the systemic inflammation originating from visceral adipose tissue, which is a metabolically active organ that produces a range of pro-inflammatory cytokines [44]. This low-grade systemic inflammation, a hallmark of central obesity, can exacerbate the local inflammatory processes within the lungs, potentially accelerating the decline in lung function and contributing to the pathogenesis of COPD [45]. Furthermore, central obesity imparts a significant mechanical load on the respiratory system. The accumulation of abdominal fat can impede diaphragmatic movement, reduce lung volumes (particularly functional residual capacity and expiratory reserve volume), and increase the work of breathing [46]. This mechanical disadvantage may worsen dyspnea and limit exercise capacity, especially in individuals with pre-existing airflow limitation [47].

Our study possesses a number of noteworthy strengths. Firstly, it is grounded on nationally representative data and incorporates sample weighting, thereby enhancing the generalizability of our results to the U.S. population. Additionally, we controlled for numerous covariates in the regression analyses and utilized the extensive sample size to conduct subgroup analyses. These steps enhance the robustness and credibility of our findings.

Nevertheless, our study is subject to several limitations. First and foremost, the diagnosis of COPD relied on self-report rather than on post-bronchodilator spirometry, the established gold standard. This approach is susceptible to potential misclassification arising from recall bias or inaccurate patient understanding of the diagnosis, which could in turn affect the magnitude of the observed association. Second, although we made efforts to adjust for numerous potential covariates, the extensive array of factors that influence COPD prevents us from completely disregarding the potential influence of other unmeasured or imperfectly measured confounding factors. For instance, variables such as dietary patterns and exposure to air pollution, which are known to be associated with both obesity and COPD, were not available in our dataset. Furthermore, physical activity was defined as a dichotomous variable rather than using a more granular scale like metabolic equivalent of task (MET) minutes per week, a simplification that might not fully account for the nuanced effects of varying activity levels. Third, it is important to explicitly state that our findings demonstrate an association rather than a causal relationship. The cross-sectional nature of our research restricts our ability to infer causality. Specifically, we cannot rule out the possibility of reverse causality. It is plausible that COPD itself could lead to changes in body composition and central fat accumulation through various mechanisms, such as systemic inflammation, metabolic dysregulation, or reduced physical activity due to dyspnea, which would in turn be reflected by a higher WWI. Therefore, prospective cohort studies are required to establish the temporal sequence and causal nature of this association. Fourth, a significant number of participants were excluded due to missing data on key variables. This could introduce selection bias if the characteristics of the excluded individuals differ systematically from those included in our final analysis, potentially limiting the generalizability of our findings. However, we have adhered to the analytical guidelines provided by the NCHS by incorporating sample weights into all analyses to account for the complex sampling design and non-response, which helps to mitigate this potential bias.

Conclusions

In conclusion, the present study has identified a noteworthy positive association between WWI and COPD, with consistent findings across various population characteristics. These findings emphasize the potential of WWI as a significant independent factor associated with the odds of having COPD. This suggests that managing central obesity, as measured by WWI, may be an important consideration in strategies aimed at mitigating the burden of COPD. Nonetheless, further basic and prospective research is required to validate these findings.