383 / 412
Figures 1 and 2 provide graphical representations of the
linear regression analyses.
Linear regression models also yielded significant
associations for gender, with males having increased size
on both tracheal CT AP diameter and width measure-
ments on CT imaging compared to females. After con-
trolling for BMI and age, there was an increase in CT
AP diameter and CT width for males of 3.34 and
2.59 mm, respectively (Table IV).
We compared each anthropomorphic variable for the
strongest effect by using standardized regression coeffi-
cients. These were estimated from a linear regression
model after the risk factors had been rescaled. Among
males, the strongest predictor of CT AP diameter and air-
way area was height, and for CT width it was BMI.
Among females, the strongest predictor of CT AP diame-
ter and airway area was BMI, and for tracheal width it
was height. However, none of the variables was signifi-
cant. Therefore, no further modeling was performed.
Among both males and females, airway area corre-
lated directly with height. Those findings are shown in
Figure 3.
DISCUSSION
Our results support BMI to be inversely related to
tracheal width on CT imaging. As shown in Figure 1, for
every 1 kg/m
2
increase in BMI, the CT width decreased
by 0.05 mm (
P
5
.0389). All anthropomorphic measure-
ments showed a trend for decreased airway dimensions
(area, AP diameter, width) with increasing BMI based
on linear regression of scatter plots. This would suggest
consistency with the trend we have clinically observed
in smaller tracheal airway sizes in patients noted to
have much larger body habitus. We speculate that this
could be a secondary effect from numerous factors, one
being increased pressure on the trachea due to increased
adiposity in these patients.
Animal studies have revealed a relationship
between the natural caudal traction of the trachea by
the thoracic contents and airway patency. Prior research
has shown there is an influence of thoracic volumes on
upper airway obstruction and compression.
11–13
Studies
have shown that obese patients have problems with
lower and upper airway compression due to increased
weight and adiposity.
14,15
Specifically, abdominal obesity
is suggested to negatively influence upper airway func-
tion during sleep.
14
It is believed that increased abdomi-
nal adiposity causes diaphragmatic compression of
intrathoracic contents, which results in their cephalic
deviation. As a result, the natural caudal traction of
intrathoracic contents via the trachea is reduced and
thus increases the distensibility of the airway. This is
Fig. 1. Airway computed tomography (CT) width (mm) versus
body mass index (BMI). The CT width on the y-axis is measured
in millimeters. BMI on the x-axis is measured using weight in kilo-
grams and height in meters and is calculated by dividing the sub-
ject’s weight by the square of his/her height (kg/m
2
). The white
circle represents one individual male subject. The solid square
represents one individual female subject. The solid line represents
the linear regression analysis for all male subjects. The dotted line
represents the linear regression analysis for all female subjects.
[Color figure can be viewed in the online issue, which is available
at
www.laryngoscope.com.]
Fig. 2. Airway area (mm
2
) versus body mass index (BMI). The air-
way area on the y-axis is measured in mm
2
. BMI on the x-axis is
measured using weight in kilograms and height in meters and is
calculated by dividing the subject’s weight by the square of his/
her height (kg/m
2
). The white circle represents one individual male
subject. The solid square represents one individual female subject.
The solid line represents the linear regression analysis for all male
subjects. The dotted line represents the linear regression analysis
for all female subjects. [Color figure can be viewed in the online
issue, which is available at
www.laryngoscope.com.]
TABLE IV.
Linear Regression Results.
Variable
CT AP, mm
CT Width, mm
Estimate (SE)
P
Value
Estimate (SE)
P
Value
Male
3.34 (0.78)
<
.0001
2.59 (0.61)
<
.0001
BMI
2
0.02 (0.03)
.5040
2
0.05 (0.025)
.0389
Age
0.03 (0.02)
.2208
2
0.002 (0.019)
.9359
Estimate is the score on change in millimeters for a given variable
based on the linear regression best fit line from the scatter plot.
AP
5
anterior-posterior diameter;
BMI
5
body mass index;
SE
5
standard error; CT
5
computed tomography.
Laryngoscope 125: May 2015
D’Anza et al.: BMI and Tracheal Airway Size
158