DOI: https://doi.org/10.21203/rs.3.rs-2422674/v1
The relationship between weight-adjusted-waist index (WWI, a newly developed obesity index) and peripheral arterial disease (PAD) is unclear. We aimed to explore the association between WWI and the prevalence of PAD in US adults. A total of 7,344 participants (males: 50.60%; females: 49.40%) from the 1999–2004 National Health and Nutrition Examination Survey (NHANES) were included in this study. WWI was calculated as waist circumference (WC) divided by the square root of weight. PAD was defined as an ankle-brachial index < 0.90 in either leg. The prevalence of PAD was 7.84%, which was respectively 3.72%, 7.23%, and 12.58% in WWI tertiles 1–3 (P < 0.001). WWI was positively associated with an elevated likelihood of PAD (OR = 1.25, 95% CI: 1.06–1.48), and the association was robust in stratified subgroups (all P for trend > 0.05). For male participants, there was a nearly linear relationship between WWI and PAD (OR = 1.35, 95% CI: 1.01–1.82). However, non-linear positive relationships were detected in females with an inflection point of 10.98 cm/√kg. A positive association was observed on the left of the inflection point (OR = 2.71, 95% CI: 1.27–5.78), while the association on the right was of no statistical significance (OR = 1.01, 95% CI: 0.77–1.33). In summary, WWI was significantly associated with an increased likelihood of PAD in US adults, with a differential association between males and females.
Peripheral artery disease (PAD) is a growing public health challenge characterized by atherosclerotic narrowing and occlusion of the peripheral arteries 1. It was reported in 2015 that more than 200 million people aged ≥ 25 years worldwide suffer from PAD, with the majority occurring in low- and middle-income countries 2. Previous evidence has established that PAD is associated with an increased risk of various adverse clinical outcomes, such as ischemic stroke, acute myocardial infarction (AMI), and non-traumatic amputation 3, 4. In a prospective cohort analysis, Criqui et al. discovered that PAD increased the mortality from cardiovascular disease by 5.9-fold and that of coronary artery disease by 6.6-fold 5. Hence, screening for the presence of PAD and implementing effective interventions are critical to reducing the burden of this condition.
Obesity, a state of abnormal fat accumulation induced by an impaired energy balance, has been recognized as a well-established and modifiable risk factor for atherosclerosis 6, 7. Indeed, much evidence indicates that obesity—assessed by body mass index (BMI)—is associated with a significantly increased risk of PAD 8–10. However, in recent years, an increasing number of studies have observed the “obesity survival paradox”, in which obese PAD patients exhibit better survival than their normal-weight counterparts 11–13. This may be partly due to the fact that BMI does not differentiate between lean mass and fat mass 14. Although waist circumference (WC) has been proposed to more accurately reflect central obesity, it is also limited as a good obesity-related index due to its high correlation with BMI 15.
In this context, some investigators have proposed a new anthropometric index, the weight-adjusted-waist index (WWI), which standardizes WC for weight 16. Thus, WWI may combine the advantages of WC but weaken the correlation with BMI. Indeed, several studies have consistently revealed that WWI is a better predictor of albuminuria, hypertension (HTN), and cardiovascular death than traditional indicators 17–19. Unfortunately, no previous research has examined the effect of WWI on PAD. Therefore, our study sought to investigate the association between WWI and PAD in adults using data from the National Health and Nutrition Examination Survey (NHANES).
Our cross-sectional study used data from NHANES, a nationwide survey implemented by the National Center of Health Statistics (NCHS), to evaluate the health and nutritional status of non-institutionalized civilians in the United States. The NHANES study design and procedures were previously explained in detail 20. The demographics and health-related information were obtained from the household interview questionnaires, while anthropometric and laboratory measurements were collected in the Mobile Examination Center. This observational study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.
The sample population was recruited from the 1999–2004 NHANES cycle. In total, 31,126 participants were involved in the interviews, in which lower extremity disease screening was assessed only in people aged ≥ 40 years. We first included 7,571 participants with complete ankle-brachial index (ABI) information. After the exclusion of participants with ABI values > 1.4 (n = 113, reflecting non-compressible arteries) or missing data on body weight (n = 41) and WC (n = 73), 7,344 eligible participants were included in the final analysis (Figure S1 in Supplementary Information).
WC (cm) divided by the square root of body weight (kg) was used to calculate WWI (cm/√kg), which was designed as an exposure variable in this study. Briefly, WC was measured to the nearest 0.1 cm at the uppermost lateral border of the ilium using a retractable steel measuring tape, and body weight was measured to the nearest 0.1 kg using a digital weight scale. The full procedure, including the protocols, equipment, and quality control, was available at the link below: https://wwwn.cdc.gov/nchs/nhanes/ContinuousNhanes/ Manuals.aspx?BeginYear = 1999.
In this analysis, PAD was designed as an outcome variable. Subjects lie supine on the exam table during the examination. According to the standard operation protocol, systolic blood pressure (SBP) was determined using an 8-MHz Doppler probe in the right arm (brachial artery) and both ankles (posterior tibial arteries). Then ABI was calculated by dividing the mean ankle SBP by the mean brachial SBP on the same side. Of note, those individuals aged 40–59 years measured SBP twice at each site and those aged 60 years and older measured SBP once. For participants who had conditions that interfered with readings in the right arm, the ABI in both legs was calculated using the left arm. Finally, the presence of PAD was defined as an ABI < 0.90 in either leg 21.
Demographic and lifestyle variables included gender, age, race/ethnicity, educational level, smoking status (yes/no), and alcohol intake (yes/no). Smoking status was assessed by “Smoked at least 100 cigarettes in life” and alcohol intake was assessed by “Had at least 12 alcohol drinks per year”. The anthropometric covariates included BMI, SBP, and diastolic blood pressure (DBP). BMI was defined as weight in kilograms divided by height in meters squared. Moreover, laboratory examinations also have been included, such as total cholesterol (TC), triglycerides (TG), total bilirubin (TBiL), hemoglobin A1c (HbA1c), aspartate transaminase (AST), alanine transaminase (ALT), serum creatinine (Scr), and serum uric acid (SUA). Besides, the health conditions, including HTN (yes/no), diabetes mellitus (DM) (yes/no), and coronary heart disease (CHD) (yes/no), were collected based on the self-report of subjects.
All analyses were performed using R version 4.0.3 (www.R-project.org) and EmpowerStates (www.empowerstats.com). Data were presented as mean ± standard deviation (SD) or median (interquartile range) for continuous variables, and frequency (percentage) for categorical variables. The normality distribution was determined by the Shapiro-Wilk test. Differences between subgroups (WWI tertiles) were detected using one-way analysis of variance (ANOVA) tests for continuous variables and Chi-square tests for categorical variables. Multivariate logistic regression models were performed to calculate the odds ratios (ORs) and 95% confidence intervals (CIs) for PAD and WWI. We further used a generalized additive model (GAM) and smooth curve fittings (penalized spline method) to address the non-linearity of WWI with PAD. Besides, subgroup analyses were also conducted using stratified logistic regression models with stratified factors including gender, age, race, education level, BMI, HTN, DM, and CHD. In addition, a receiver operating characteristic (ROC) curve was employed to analyze the predictive value of different obesity-related indices for PAD. A two-sided P < 0.05 was considered statistically significant.
All methods were performed in line with the Declaration of Helsinki. The NHANES experimental protocols were approved by the NCHS institutional review board (Protocol #98 − 12), with all individuals providing written informed consent.
Baseline characteristics
The baseline characteristics of the study population stratified by WWI tertiles are shown in Table 1. A total of 7,344 participants (males: 50.60%; females: 49.40%) with an average age of 60.18 ± 13.03 years were included in this study. The mean WWI was 11.16 ± 0.78 cm/√kg, and the ranges of tertiles 1-3 were 8.36- 10.81, 10.81-11.49, and 11.49-15.52 cm/√kg, respectively. Moreover, the prevalence of PAD was 7.84% (576 participants), which was respectively 3.72% (91 participants), 7.23% (177 participants), and 12.58% (308 participants) in WWI tertiles 1-3 (P <0.001). Individuals in the highest WWI tertile were predominated by older participants, and had higher BMI, WC, SBP, HbA1c, TC, TG, and SUA; higher prevalence of HTN, DM, and CHD; less alcohol intake; and lower educational level and DBP values (all P <0.001).
Table 1. Baseline characteristics of the study population according to weight-adjusted-waist index tertiles.
Characteristics |
Overall |
Tertile categories of WWI (cm/√kg) |
P value |
||
T1 (8.36-10.81) |
T2 (10.81-11.49) |
T3 (11.49-15.52) |
|||
Participants |
7344 |
2448 |
2448 |
2448 |
|
Age (years) |
60.18 ± 13.03 |
54.06 ± 11.55 |
60.15 ± 12.30 |
66.32 ± 12.23 |
<0.001 |
Gender (%) |
|
|
|
|
<0.001 |
Male |
3716 (50.60%) |
1278 (52.21%) |
1373 (56.09%) |
1065 (43.50%) |
|
Female |
3628 (49.40%) |
1170 (47.79%) |
1075 (43.91%) |
1383 (56.50%) |
|
Race (%) |
|
|
|
|
<0.001 |
Mexican American |
1540 (20.97%) |
350 (14.30%) |
523 (21.36%) |
667 (27.25%) |
|
Other Hispanic |
285 (3.88%) |
68 (2.78%) |
115 (4.70%) |
102 (4.17%) |
|
Non-Hispanic White |
3968 (54.03%) |
1343 (54.86%) |
1298 (53.02%) |
1327 (54.21%) |
|
Non-Hispanic Black |
1326 (18.06%) |
620 (25.33%) |
430 (17.57%) |
276 (11.27%) |
|
Other Race |
225 (3.06%) |
67 (2.74%) |
82 (3.35%) |
76 (3.10%) |
|
Educational level (%) |
|
|
|
|
<0.001 |
Less than high school |
2468 (33.67%) |
572 (23.42%) |
788 (32.26%) |
1108 (45.32%) |
|
High school or GED |
1726 (23.55%) |
543 (22.24%) |
609 (24.93%) |
574 (23.48%) |
|
More than high school |
3136 (42.78%) |
1327 (54.34%) |
1046 (42.82%) |
763 (31.21%) |
|
BMI (kg/m2) |
28.40 ± 5.58 |
26.03 ± 4.80 |
28.64 ± 5.24 |
30.57 ± 5.72 |
<0.001 |
WC (cm) |
98.97 ± 13.87 |
89.31 ± 10.81 |
99.87 ± 11.22 |
107.71 ± 12.80 |
<0.001 |
WWI (cm/√kg) |
11.16 ± 0.78 |
10.32 ± 0.40 |
11.14 ± 0.20 |
12.01 ± 0.43 |
<0.001 |
SBP (mmHg) |
133.83 ± 22.28 |
127.57 ± 20.49 |
133.83 ± 21.59 |
140.15 ± 22.93 |
<0.001 |
DBP (mmHg) |
72.96 ± 14.37 |
74.37 ± 13.02 |
73.24 ± 13.76 |
71.27 ± 16.04 |
<0.001 |
Smoking status (at least 100 cigarettes in life) (n, %) |
3948 (53.77%) |
1288 (52.61%) |
1325 (54.13%) |
1335 (54.56%) |
0.453 |
Alcohol intake (at least 12 alcohol drinks per year) (n, %) |
4809 (66.99%) |
1732 (72.14%) |
1630 (68.09%) |
1447 (60.70%) |
<0.001 |
Hemoglobin A1c (%) |
5.77 ± 1.14 |
5.51 ± 0.89 |
5.75 ± 1.06 |
6.06 ± 1.34 |
<0.001 |
TBiL (umol/L) |
11.93 ± 4.86 |
12.25 ± 5.02 |
12.09 ± 5.04 |
11.44 ± 4.45 |
<0.001 |
AST (U/L) |
26.00 ± 25.81 |
26.36 ± 39.14 |
26.22 ± 16.84 |
25.42 ± 13.70 |
0.408 |
ALT (U/L) |
26.06 ± 31.12 |
26.04 ± 45.26 |
27.01 ± 23.59 |
25.11 ± 17.43 |
0.111 |
Total cholesterol (mmol/L) |
5.38 ± 1.07 |
5.32 ± 1.01 |
5.41 ± 1.08 |
5.40 ± 1.10 |
0.004 |
Triglycerides (mmol/L) |
1.75 ± 1.69 |
1.50 ± 1.74 |
1.76 ± 1.53 |
1.99 ± 1.74 |
<0.001 |
Serum uric acid (µmol/L) |
327.25 ± 86.40 |
308.59 ± 82.06 |
333.89 ± 85.95 |
339.20 ± 88.00 |
<0.001 |
Serum creatinine (µmol/L) |
78.96 ± 38.69 |
77.65 ± 31.88 |
79.25 ± 40.79 |
79.99 ± 42.52 |
0.107 |
Hypertension (%) |
3075 (42.04%) |
733 (30.08%) |
1033 (42.39%) |
1309 (53.65%) |
<0.001 |
Diabetes mellitus (%) |
956 (13.02%) |
158 (6.45%) |
273 (11.15%) |
525 (21.45%) |
<0.001 |
Coronary heart disease (%) |
467 (6.36%) |
85 (3.47%) |
179 (7.31%) |
203 (8.29%) |
<0.001 |
Peripheral arterial disease (%) |
576 (7.84%) |
91 (3.72%) |
177 (7.23%) |
308 (12.58%) |
<0.001 |
Abbreviations: WWI, weight-adjusted-waist index; BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; TBiL, total bilirubin; AST, aspartate transaminase; ALT, alanine transaminase.
Associations between WWI and PAD
As shown in Table 2, elevated WWI was associated with an increased likelihood of PAD. After full adjustment (Model 3), every one unit increase in WWI was associated with 25% increased odds of PAD (OR = 1.25; 95% CI: 1.06-1.48). We further transformed WWI into categorical variables for sensitivity analysis. The adjusted OR was 1.53 (95% CI: 1.10-2.12) for Tertile 2 and 1.77 (95% CI: 1.26-2.49) for Tertile 3, indicating a stable positive association between WWI and increased risk of PAD with statistical significance (P for trend = 0.002).
Table 2. The association between weight-adjusted-waist index and peripheral arterial disease in total samples.
WWI (cm/√kg) |
Events (%) |
PAD, OR (95%CI), P value |
||
Model 1 |
Model 2 |
Model 3 |
||
Per 1cm/√kg increase |
576 (7.84) |
1.94 (1.74, 2.17), <0.001 |
1.39 (1.19, 1.63), <0.001 |
1.25 (1.06, 1.48), 0.009 |
Tertiles |
|
|
|
|
T1 (8.36-10.81) |
91 (3.72) |
Reference |
Reference |
Reference |
T2 (10.81-11.49) |
177 (7.23) |
2.02 (1.56, 2.62), <0.001 |
1.62 (1.20, 2.21), 0.002 |
1.53 (1.10, 2.12), 0.011 |
T3 (11.49-15.52) |
308 (12.58) |
3.73 (2.93, 4.74), <0.001 |
2.09 (1.52, 2.88), <0.001 |
1.77 (1.26, 2.49), 0.001 |
P for trend |
|
<0.001 |
<0.001 |
0.002 |
Abbreviations: WWI, weight-adjusted-waist index; PAD, peripheral arterial disease; 95% CI: 95% confidence interval; OR: odds ratio;
Model 1: No covariates were adjusted.
Model 2: Adjusted for gender, age, race, education level, body mass index, systolic and diastolic blood pressure.
Model 3: Adjusted for model 2 + smoking status, alcohol intake, hemoglobin A1c, total bilirubin, aspartate transaminase, alanine transaminase, total cholesterol, triglycerides, serum creatinine, serum uric acid, hypertension, diabetes mellitus, and coronary heart disease.
Besides, the smooth curve fitting indicated that there was a nearly linear relationship between WWI and the risk of PAD in the total population (Figure 1).
Subgroup analysis
Subgroup analyses were conducted to explore the robustness of the association between WWI and PAD in different population settings. None of the stratifications, including gender (male and female), age (<60 years and ≥60 years), race (Mexican American, other Hispanic, non-Hispanic black, non-Hispanic white, and other races), education level (< high school, high school, and > high school), BMI (<30 kg/m2 and ≥30 kg/m2), HTN (yes and no), DM (yes and no), and CHD (yes and no), significantly modified the association between WWI and PAD prevalence (all P for interaction > 0.05). The results indicated that the positive WWI-PAD association was stable in stratified subgroups and could be appropriate for various populations (Figure 2).
Non-linear positive association of WWI and PAD in females
Given the difference in body composition between males and females, analyses stratified by sex were further conducted. As shown in Table S1 (Supplementary Information), a significantly positive association between WWI and PAD prevalence was observed in male participants (Model 3: OR = 1.35, 95% CI: 1.01-1.82). Although there were similar trends in female participants, the difference was not statistically significant (Model 3: OR = 1.23, 95% CI: 0.99-1.52).
We then used GAM and smooth curve fitting to evaluate the non-linearity for different sexes. The results show that there was a non-linear relationship between WWI and the prevalence of PAD in female participants, with a log-likelihood ratio test of 0.021 (Table 3; Figure 3). The inflection point was determined to be 10.98 cm/√kg using the two-piecewise linear regression model. On the left of the inflection point, a positive association between WWI and PAD was observed, with the adjusted OR being 2.71 (95% CI: 1.27-5.78). Nevertheless, no relationship with statistical significance was observed on the right of the inflection point (OR = 1.01, 95% CI: 0.77-1.33). Unlike females, a nearly linear relationship was detected in males (OR = 1.35, 95% CI: 1.01-1.82), with a log-likelihood ratio test of 0.317 (Table 3; Figure 3).
Table 3. Threshold effect analysis of WWI on PAD using a two-piecewise linear regression model.
WWI (cm/√kg) |
*Adjusted OR (95% CI), P value |
||
Male |
Female |
All participants |
|
Fitting by standard linear model |
1.35 (1.01, 1.82), 0.043 |
1.23 (0.99, 1.52), 0.058 |
1.25 (1.06, 1.48), 0.009 |
Fitting by two-piecewise linear model |
|
|
|
Inflection point (K) |
11.44 |
10.98 |
11.64 |
< K Effect size |
0.77 (0.26, 2.30), 0.645 |
2.71 (1.27, 5.78), 0.010 |
1.50 (1.14, 1.97), 0.004 |
> K Effect size |
1.44 (1.05, 1.98), 0.024 |
1.01 (0.77, 1.33), 0.930 |
1.00 (0.72, 1.37), 0.986 |
Log likelihood ratio test P-value |
0.317 |
0.021 |
0.093 |
Abbreviations: WWI, weight-adjusted-waist index; PAD, peripheral arterial disease; 95% CI: 95% confidence interval; OR: odds ratio.
*Adjusted for gender (only for all participants), age, race, education level, body mass index, systolic and diastolic blood pressure, smoking status, alcohol intake, hemoglobin A1c, total bilirubin, aspartate transaminase, alanine transaminase, total cholesterol, triglycerides, serum creatinine, serum uric acid, hypertension, diabetes mellitus, and coronary heart disease.
WWI predicts PAD better than BMI and WC
The ROC curves were used to evaluate the discriminative ability of different indicators for PAD. As shown in Table 4 and Figure S2 (Supplementary Information), WWI showed the highest area under the curve (AUC) in both males (0.647; 95% CI: 0.615-0.680) and females (0.654; 95% CI: 0.623-0.685), with cut-off values of 11.12 and 11.36 cm/√kg, respectively.
Table 4. Comparison of the ability of different obesity-related indices to predict peripheral artery disease.
Test |
AUC (95% CI) |
Cut-off |
Specificity |
Sensitivity |
Youden index |
Males |
|
|
|
|
|
BMI |
0.567 (0.532, 0.602) |
28.55 |
0.424 |
0.703 |
0.127 |
WC |
0.501 (0.466, 0.536) |
112.75 |
0.819 |
0.208 |
0.027 |
WWI |
0.647 (0.615, 0.680) |
11.12 |
0.554 |
0.698 |
0.252 |
Females |
|
|
|
|
|
BMI |
0.505 (0.470, 0.540) |
31.16 |
0.700 |
0.333 |
0.033 |
WC |
0.559 (0.526, 0.592) |
97.05 |
0.564 |
0.559 |
0.123 |
WWI |
0.654 (0.623, 0.685) |
11.36 |
0.588 |
0.653 |
0.241 |
Abbreviations: WWI, weight-adjusted-waist index; BMI, body mass index; WC, waist circumference; AUC, area under the curve; 95% CI, 95% confidence interval.
In this nationally representative cross-sectional study, we observed that participants with increased WWI were significantly associated with an increased likelihood of PAD. The relationship was independent of multiple potential covariates including demographics, laboratory measurements, cardiovascular risk factors, and others. Subgroup analysis indicated that the WWI-PAD association was stable in different populations. Furthermore, non-linear positive relationships between WWI and the prevalence of PAD were detected in females, and the different correlations were found on the left and right sides of the inflection point (10.98 cm/√kg). Besides, WWI was superior to BMI and WC in predicting PAD, suggesting that the management of obesity evaluated by WWI may reduce the risk of PAD.
To our knowledge, this is the first report assessing the association between WWI and PAD in a nationally representative population. As a new simple adiposity index, WWI has been investigated in various fields, especially related to cardiometabolic diseases 17-19, 22. In a cohort study involving 465,629 individuals, Park et al. found that WWI was positively associated with cardiometabolic morbidity and cardiovascular death 16. Moreover, when compared to the lowest WWI subgroup (< 9.94 cm/√kg), Li et al. discovered that the highest WWI subgroup (≥ 10.91 cm/√kg) was associated with 50% increased risk of HTN 18. In addition, a cross-sectional study of 14,078 hypertensive patients by Zhao et al. aimed to investigate the relationship between WWI and hyperuricemia 23. The results show that per one unit increase in WWI (Model 3), the risk of hyperuricemia increased by 37% (95% CI: 1.25-1.49) in males and by 35% (95% CI: 1.26-1.45) in females. In this study, we first reported the positive association and gender differences between WWI and the prevalence of PAD, and the relationship was stable in different subgroups evaluated by the stratified analyses.
Previous studies have shown that the traditional obesity-related index, BMI, is related to PAD. A population-based study involving 11,477 Chinese community-dwelling adults suggested that each standard deviation (3.60 kg/m2) increase in BMI was correlated with 23% (95% CI: 1.13-1.33) incremental PAD risk. Moreover, the categorical analysis showed that compared with the lowest quartile of BMI, the highest quartile was associated with 31% (95% CI: 1.03-1.67) increased PAD risk (P for trend = 0.008) 9. In the China Hypertension Registry study, Li et al. found that BMI was positively associated with the risk of PAD (OR = 1.52; 95% CI: 1.52-1.93) in those with BMI greater than 25.7 kg/m2 24. However, some studies indicate an inverse association between BMI and the prevalence of PAD 25. A meta-analysis including 5,729,272 individuals further demonstrated that higher BMI values were associated with a lower mortality risk in PAD patients 13. Because BMI does not distinguish between lean mass and fat mass, this phenomenon may be due to the limitations of BMI and the existence of the “obesity survival paradox”. Although WC was proposed as a surrogate measure for the indirect evaluation of visceral adipose tissue that more accurately reflects adverse metabolic profiles, it is also limited by its high correlation with BMI. In this context, some nontraditional obesity indicators were also performed to explore the association with PAD. A cross-sectional study of 1,872 patients with diabetes mellitus by Wung et al. showed that the body roundness index (BRI) was significantly associated with peripheral artery occlusive disease (P = 0.021), and the AUC was 0.630 (95 CI%: 0.567-0.692) 26. Another cross-sectional study using data from the China H-type Hypertension Registry demonstrated that visceral adiposity index (VAI) was positively associated with PAD in normal-weight adults with HTN 27. However, most of these nontraditional indicators are based on relatively complicated mathematical models, and therefore routine examinations in the general population would be limited.
Notably, WWI is a new and simple indicator based on the formula [In(WC) = β0 + β1 In(weight) + ε] that has the potential to combine the advantages of WC while weakening the correlation with BMI, enabling it to assess both fat and muscle mass components 16. Using a cross-sectional sample of 602 participants aged ≥ 65 years from the Ansan Geriatric Study, Kim et al. found that WWI was positively correlated with total abdominal fat area but negatively correlated with appendicular skeletal muscle mass 28. Moreover, several prospective cohort studies further confirmed that WWI rather than BMI and WC better predicted the risk of all-cause and cardiovascular death 16, 19. Consistent with our current results, WWI was superior to BMI and WC in predicting PAD, indicating that WWI might be a better predictor of PAD than other obesity parameters. These findings suggest that WWI may be a comprehensive and superior indicator of obesity and could be used to predict obesity-related disorders in clinical settings.
Given that WWI has been shown to more accurately estimate total body fat percentage, the potential mechanism of the positive WWI-PAD association could be related to metabolic abnormalities. There is evidence that even normal-weight individuals are metabolically obese 29. Moreover, even with a similar degree of obesity, WWI was significantly higher in metabolically unhealthy groups than in metabolically healthy groups 30. Hence, elevated WWI may reflect a state of metabolic disorder with excessive adipose tissue accumulation. The dysfunction of adipose tissue secretes various inflammatory cytokines and adipocytokines (such as interleukin-6 and resistin), leading to insulin resistance, endothelial dysfunction, and inflammation response, which are correlated with an increased risk of HTN, DM, and PAD 31-33. Besides, we observed gender differences in WWI-PAD association, which may be explained by the specific fat distribution. As previous studies have shown that total abdominal fat and visceral fat were both significantly lower in females than in males 34. Other potential reasons are that estrogen may slow down the adverse effects of adipose tissue on the progression of atherosclerosis 35.
Our study has the strengths of a large population-based sample size, rigorous study protocols and quality controls, and data available on many important covariates by integrating the NHANES survey. By adjusting for multiple potential covariates, we confirmed the robustness of the results. Furthermore, we explored the dose-response relationship between WWI and PAD and found significant gender differences, which are very easily overlooked in clinical practice. Despite that, the limitations of this study should also be declared. First, although we adjusted for many important variables, we could not completely rule out the influence of other potential confounders, such as medications and hormones, which may affect the progression of atherosclerosis. Second, the cross-sectional nature of this investigation made it impossible to determine the causal relationship between WWI and PAD. Third, all participants were recruited from a single country, thus limiting the ability to generalize our findings to the general population.
This cross-sectional study suggested that WWI levels were significantly associated with an increased likelihood of PAD in United States adults, with a differential association between males and females. WWI was superior to BMI and WC in predicting PAD, indicating that the management of obesity evaluated by WWI may reduce the risk of PAD. Despite that, further large-scale longitudinal studies are still needed to verify our findings.
Acknowledgments
We thank all National Health and Nutrition Examination Survey (NHANES) participants and staff for their valuable efforts and contributions.
Author contributions
F.X. and Y.W. conceived and designed the study. R.G., H.F., and J.X. contributed to data collection and statistical analysis. F.X. drafted the manuscript. Y.W. had primary responsibility for the final content of the manuscript. All authors interpreted the results, and reviewed and approved the manuscript.
Funding
This work was supported by the National Key Research and Development Plan of China (No. 20212BBG71004), the National Natural Science Foundation of China (No. 82160085), the Jiangxi Provincial Department of Science and Technology Foundation (No. 20181BCB24013), the Natural science funding (No. 20202BAB206005), and the Jiangxi Provincial Natural Science Foundation (20192BAB205012).
Competing interests
The authors declare no competing interests.
Additional information
Correspondence and requests for materials should be addressed to Y.W.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Data availability
Publicly available datasets were analyzed in this study. These data can be found at: www.cdc.gov/nchs/nhanes/index.htm.