Ectopic fat accumulation in patients with COPD: an ECLIPSE substudy

Figures & data

Figure 1 Body composition analysis and differences in fat distribution.

Notes: Examples of a CT slice at L2–L3 before (A) and after segmentation (B). An example is also provided for two patients with COPD with same BMI and similar waist circumference but marked difference in body composition, with one patient (C) having more VAT (shown in red) than the other patient (D). After segmentation, SCAT was identified in green, VAT in red and MT in pink. CSAs of tissues (cm²) were computed on the segmented structures using an attenuation range of −190 to −30 HUs for AT and −29 to 130 HUs for MT. The mean attenuation value in HU of each structure was also quantified.
Abbreviations: CT, computed tomography; BMI, body mass index; SCAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; MT, muscle tissue; CSA, cross-sectional area; HU, Hounsfield unit; AT, adipose tissue.

Table 1 Baseline characteristics by group

Variables	COPD (n=511)	Controls (n=74)	P-value
Age (years)	63.8±6.9	56.8±10.0	<0.0001
BMI (kg/m^2)	24.4±5.0	25.3±4.3	0.1761
Sex
Male, n (%)	314 (61)	30 (41)	0.0007
Female, n (%)	197 (39)	44 (59)
Smoking status
Current smokers, n (%)	201 (39)	20 (27)	<0.0001
Former smokers, n (%)	310 (61)	17 (23)
Never-smokers, n (%)	0 (0)	37 (50)
Pack/year	48.7±1.1	14.9±4.4	<0.0001
Pulmonary function
FEV1 (% predicted value)	40.7±15.3	108.3±13.2	<0.0001
FVC (L)	2.71±0.9	4.07±0.9	<0.0001
FEV1/FVC	0.41±0.01	0.79±0.01	<0.0001
FEV1 decline (L/year)	0.052±0.023	NA	NA
Clinical variables
SGRQ total score	48.4±17.2	9.5±9.7	<0.0001
6MWD (m)	366.3±5.7	NA	NA
Exacerbation number (number/years)	1.3±0.1	0 (0)	<0.0001
Mortality, n (%)	49 (9.5)	1 (1.3)	0.02
Mean survival (days)	1,016±6	1,060±5	<0.01
Comorbidities
Cardiovascular†, n (%)	255 (50)	1 (1)	0.0078
Hypertension, n (%)	175 (34)	9 (12)	0.0001
CAC score, Agatston units	540±45	251±120	0.0005
Diabetes, n (%)	31 (6)	1 (1)	0.1068
Reflux/heartburn, n (%)	110 (22)	15 (20)	0.8802

Notes: Data shown as mean ± SE or n (%).

† Includes angina, heart attack, myocardial infarction, stroke and heart failure.

Abbreviations: BMI, body mass index; FEV₁, forced expiratory volume in 1 second; FVC, forced vital capacity; NA, not applicable; SGRQ, St George’s Respiratory Questionnaire; 6MWD, 6-minute walking distance; CAC, coronary artery calcification; SE, standard error.

Table 2 Internal validity and external validity of the CT image analysis

Variables	Internal validity		External validity
	Pearson’s r	Intraclass r	Pearson’s r	Intraclass r
VAT surface	1.00	1.00	1.00	1.00
MT surface	0.99	0.99	0.99	0.99
MT attenuation	1.00	1.00	0.95	0.95
SCAT	1.00	1.00	0.99	0.99

Abbreviations: CT, computed tomography; VAT, visceral adipose tissue; MT, muscle tissue; SCAT, subcutaneous adipose tissue.

Figure 2 Group mean values ± SD for BMI (A), CSA of VAT (B), MT attenuation (C), CSA of MT (D) and CSA of SCAT (E) in patients with COPD (gray bars) and in controls (black bars).

Note: Patients with COPD were further classified according to the GOLD spirometric classification system, with a greater score indicating worse airflow obstruction.
Abbreviations: BMI, body mass index; CSA, cross-sectional area; VAT, visceral adipose tissue; MT, muscle tissue; SCAT, subcutaneous adipose tissue; GOLD, Global Initiative for Chronic Obstructive Lung Disease; HU, Hounsfield unit; SD, standard deviation.

Figure 3 Trend analysis between CT body composition parameters and clinical outcomes in COPD.

Notes: Trend analysis between CSA of VAT (A), MT attenuation (B), and CSA of MT (C), and the 6MWD, exacerbation rate, SGRQ and FEV₁ decline. For each body composition parameter, the panels report the odds ratio (mean and 95% CI) of having a worse clinical outcome with the deterioration of the body composition parameter from quartile 1 (reference) to quartiles 2, 3 and 4. The P-values indicate the statistical significance for the trend of the association.

Abbreviations: CT, computed tomography; CSA, cross-sectional area; VAT, visceral adipose tissue; MT, muscle tissue; 6MWD, 6-minute walking distance; SGRQ, St George’s Respiratory Questionnaire; FEV₁, forced expiratory volume in 1 second; CI, confidence interval.

Figure 4 Logistic regression models indicating significant correlations between CT body composition parameters and the proportion of comorbidities for diabetes (A), gastroesophageal reflux (B), and CVD (C) at baseline.

Note: Continuous lines were drawn from the mean regression coefficients with the dashed lines representing the 95% CI of the relationships.
Abbreviations: CT, computed tomography; CI, confidence interval; VAT, visceral adipose tissue; CSA, cross-sectional area; CVD, cardiovascular disease; MT, muscle tissue; HU, Hounsfield unit.