Effect of truncal obesity on airway resistance

Background and objective: Truncal obesity can influence respiratory mechanisms regardless of BMI. This study evaluated the impact of truncal obesity on both spirometry and impulse oscillometry (IOS). Patients and methods: The study included 102 patients recruited from the Outpatient Department of Ain Shams University Hospital. All patients were subjected to clinical examination, anthropometric measurements, IOS, and spirometry. According to BMI, the patients were categorized into the obese group (71 patients) and the nonobese group (31 patients). The obese group was subclassified into overweight, obese, and morbidly obese groups. The obese subgroup was reclassified according to waist to hip ratio (WHR) into the truncal obese group (41 patients) and the nontruncal obese group (30 patients). The data collected were comparatively analyzed using either the independent-samples t-test or one-way analysis of variance. Pearson′s correlation coefficient was used to study the correlations between data. Multiple linear regression analysis was carried out to examine the simultaneous influences of anthropometric measures on the IOS data. Results: There were significant differences in IOS and spirometry data between the obese and nonobese groups as well as among the four groups. Regarding WHR, there were significant correlations with IOS data and negative correlations with spirometry data except maximal mid expiratory flow. Using multiple linear regression analysis, it was revealed that WHR was the most powerful predictor for R5. BMI followed by WHR was the most powerful predictor for R5-R20. Conclusion: Truncal obesity significantly affects airway resistance, as evaluated by spirometry and IOS.


Introduction
Th e study was approved by the review board of Pulmonary Medicine Department, Ain Shams University, and informed consent was obtained from all patients.
All patients underwent history taking and clinical examination, anthropometric measurements including height, weight, BMI, waist, hip and neck circumferences, and WHR, IOS, and prebronchodilator and postbronchodilator spirometry.
Th e obese patients were reclassifi ed according to WHR into the truncal obese group (WHR ≥0.90 in males and ≥0.85 in females), which included 41 patients, and the nontruncal obese group (WHR <0.90 in males and <0.85 in females), which included 30 patients [6].

Spirometry
Spirometry was performed using Spirometric s ENC Flowmate (Spring Valley, New York, USA). The test was performed before and 20 min after β2-agonist inhalation (salbutamol 400 μg) by means of a metered-dose inhaler. Prebronchodilator and postbronchodilator spirometry were measured according to American Thoracic Society/European Respiratory Society standards in all patients [8]. Proximal obstruction is considered if the following conditions are fulfi lled: (1) R5 is high (>150% predicted) (i.e. in the abnormal range). (2) R20 is elevated (>150% predicted).
(3) R(f) is independent of frequency and almost horizontal (i.e. the R20 is similar to the R5). (4) X5 is completely normal, as the resonant frequency.
(3) R(f) is frequency dependent and becomes lower at higher frequencies. (4) X5 is reduced to the abnormal range, and the resonant frequency (f res ) is shifted to the right (i.e. to higher frequencies) [11].

Data analysis
Data were compared using analysis of variance, followed by post-hoc multiple comparisons by the Bonferroni method.
Simple correlations between data were tested using Pearson's correlation coeffi cient.
Multiple linear regression test was performed to examine the concurrent eff ects of anthropometric measures including calculated BMI, WHR, and measured neck circumference, with the IOS data R5, R2 0, R5−R20, and X5 as dependent variables.
Th e statistical software statistical package for the social scienc es (SPSS, version 17; SPSS Inc., Chicago, Illinois, USA) was used for statistical analysis. All tests were considered signifi cant at P value less than 0.05.  Table 1. According to BMI, the patients were categorized into two groups: the obese group (BMI ≥25), which included 71 (69.6%) patients, and the nonobese group (BMI <25), which included 31 (30.4%) patients. Th e obese and nonobese groups were compared using independent-samples t-test for IOS and spirometry data. Th ere were signifi cant diff erences between the two groups and they are outlined in Table 2.
Th e obese patients were reclassifi ed according to WHR into the truncal obese group (WHR ≥0.90 in males and ≥0.85 in females), which included 41 (57.7%) patients, and the nontruncal obese group (WHR <0.90 in males and <0.85 in females), which included 30 (42.3%) patients. Th e two groups were compared as regards the IOS and spirometry data using the independent-samples t-test. Th ese results are presented in Table 3.
Regarding WHR there were significant positive correlations with IOS data -namely, R5 (r = 0.712,  Comparison between the four groups with respect to impulse oscillometry (IOS) data using one-way analysis of variance followed by multiple comparisons with the post-hoc Bonferroni method. There were signifi cant differences between the four groups regarding total airway resistance (R5) (F = 42.47, P = 0.000), central airway resistance (R20) (F = 10.93, P = 0.000), R5−R20 (F = 26.25, P = 0.000), and X5 ( F = 11.87, P = 0.000). Simultaneous eff ects of diff erent anthropometric data including BMI, neck circumference, and WHR were analyzed by multiple linear regression analysis, with the IOS data (R5, R20, and R5−R20) as the dependent variables, and independent variables being selected by the stepwise procedure.  Fig. 8.
A third multiple regression model was tried using R20 as the dependent variable. Th e model accounted for only 18% of estimation of R20 and none of the anthropometric variables were signifi cant contributors in this estimation.

Discussion
Th e eff ect of obesity and overweight on pulmonary mechanics and airway resistance is the main concern of this study. As expected, our data showed that there were signifi cant diff erences between the two groups, obese and nonobese, as regards spirometric data Comparison between the four groups with respect to spirometry data using one-way analysis of variance followed by multiple comparisons with the post-hoc Bonferroni method. There were significant differences between the four groups regarding forced vital capacity (FVC) (F = 75.86, P = 0.000), forced expiratory volume in 1 s (FEV 1 ) (F = 14.50, P = 0.000), and FEV 1 /FVC ratio (F = 12.84, P = 0.000). There was no signifi cant difference regarding maximal mid expiratory fl o w (MMEF) (P > 0.05).

Fig. 2
There was a signifi cant positive correlation between waist to hip ratio (WHR) and total airway resistance (R5) (r = 0.712, P = 0.000) on using Pearson's correl ation coeffi cient.  including FVC, FEV 1 , and FEV 1 /FVC data except for MMEF. Th ere was signifi cant diff erence between the two groups as regards IOS data, especially R5, which gives information about the entire respiratory tract, as well as R20, R5−R20, which indicates the central and the peripheral airway resistance, and X5, the distal capacitive reactance elastic lung and thorax components aff ecting compliance and therefore in turn aff ecting small airway resistance. Th ese results were similar to those studies that proved that obesity decreases the total respiratory system compliance by two-thirds and also increases airway resistance. Moreover, these studies proved that FEV 1 is lower in obese patients compared with nonobese patients; obese patients also had their fl ow rates at 50 and 75% of exhaled vital capacity decreased. Further, airway resistance was signifi cantly greater in obese patients [12,13].
Th e results of this study showed that IOS detected peripheral airway resistance in the obese group, and diagnosing small airway disease gives it an advantage over spirometry. Th is agrees with those studies that concluded that IOS provides additional information not obtained by simple spirometry and that airfl ow resistance measurements were not well predicted by spirometry [14,15].
Th e four subgroups of normal, overweight, obese, and morbidly obese were compared with regard to IOS and spirometry data. Th ere were statistically signifi cant diff erences between the four groups with respect to all data IOS and spirometric data. Th is was previously proved by various studies: for example, the observations of Zerah et al. [16] who examined airway resistance in a There was a signifi cant positive correlation between waist to hip ratio (WHR) and central airway resistance (R20) (r = 0.351, P = 0.000) on using Pearson's correl ation coeffi cient.

Fig. 5
There was a signifi cant negative correlation between waist to hip ratio (WHR) and forced vital capacity (FVC) (r = −0.606, P = 0.000) on using Pearson's corre lation coeffi cient. Multiple linear regression model with total airway resistance (R5) as the dependent variable and the anthropometric measures BMI, neck circumference, and waist to hip ratio as independent variables selected by the stepwise procedure. The model is signifi cant (F = 57.412, P = 0.000) with a coeffi cient of de termination of 63%. group of obese patients who were subdivided into mildly obese, moderately obese, and morbidly obese groups; in their study airway resistance increased signifi cantly with the degree of obesity, and was inversely related to changes in the functional residual capacity and also to the elastic load. Obese individuals must overcome the increased airway resistance resulting from the decrease in lung volume due to obesity [16].
In this study the obese patients were reclassifi ed according to WHR into truncal obese (WHR ≥0.90 in males and ≥0.85 in females), which included 41 (57.7%) patients, and nontruncal obese (WHR<0.90 in males and <0.85 in females), which included 30 (42.3%) patients. Th e two groups were compared as regards the IOS and showed a statistically signifi cant diff erence between them with respect to R5 and R5−R20 but no signifi cant diff erence with respect to R20 and X5, thus indicating increased peripheral airway resistance. Th e primary reason is obviously due to a decrease in chest wall compliance, which results from the accumulation of fatty adipose tissues in the abdomen, diaphragm, and around the lower ribs of obese individuals. Th us, it was previously proved that the total respiratory compliance is markedly decreased in obese patients having truncal obesity compared with obese individuals without truncal obesity; this reduction is mostly due to the decreased compliance, although it may also be due to an increase in airway resistance [17].
In this study, there was an insignifi cant diff erence with respect to R20; thus, peripheral resistance is the main factor causing signifi cant increase in airway resistance in the group with truncal obesity. With regard to WHR, there were signifi cant correlations with IOS data -namely, R5, R5−R20, R20, and X5. Also there were signifi cant negative correlations between WHR and spirometry data -namely, FVC, FEV 1 , and FEV 1 / FVC ratio, but there was no signifi cant correlation as regards MMEF, which may be less sensitive than IOS in detecting airway resistance formed by the peripheral airway aff ection.
When the truncal and nontruncal obese groups were compared for spirometry data, there was a signifi cant diff erence between the two groups only for FVC.
Simultaneous eff ects of diff erent anthropometric data including BMI, neck circumference, and WHR were analyzed using multiple linear regression analysis, with the IOS data (R5, R20, and R5−R20) as the dependent variables, and independent variables selected by the stepwise procedure. Th e most powerful predictor of R5 was the WHR, whereas the most robust predictor of peripheral resistance was BMI, followed by WHR. However, the study of the simultaneous eff ects of BMI, neck circumference, and WHR on R20, which measures the R20, revealed that they accounted for only 18% of estimation of R20 and none of the anthropometric variables were signifi cant contributors in this estimation. Th is confi rms the eff ect of truncal obesity on the peripheral and consequently R5 without a direct eff ect on R20. All of the above results confi rm those reached by a previous research that concluded that truncal obesity is most likely to aff ect pulmonary volumes, without direct eff ects on pulmonary obstruction [7].
In conclusion, truncal obesity signifi cantly aff ects airway resistance, as evaluated by sp irometry and IOS.