Patient selection

Data were abstracted from the Integrated Knowledge System based on BioXM™ (Biomax Informatics AG, Munich, Germany) of 294 patients with a primary diagnosis of COPD [post-bronchodilator forced expiratory volume in 1 s (FEV₁) /forced vital capacity (FVC) ratio of ⩽0.70 and post-bronchodilator FEV₁ < 80 % predicted] who were evaluated during the initial assessment of a comprehensive pulmonary rehabilitation (PR) program at CIRO Horn (The Netherlands) between August 2014 and May 2015 [14]. Exclusion criteria were: a history of asthma, α1-antitrypsin deficiency, lung cancer, any previous lung surgery, instable inflammatory or endocrine diseases, acute myocardial infarction within the last 6 months, any known bone disease other than osteoporosis, treated malignant disease within the last 5 years and an exacerbation of COPD within the last 4 weeks. Ethical approval was not indicated because all of the tests were done as part of the routine initial assessment [14], and analyzed retrospectively. The Board of Directors of CIRO approved the use of de-identified patients’ records.

Lung function measurement

As part of the routine 3-day initial assessment of integrated health status, all patients performed post-bronchodilator spirometry (Masterlab®; Jaeger, Würzburg, Germany). FEV₁ and FVC were assessed in accordance with the latest GOLD guidelines [1]. Inspiratory vital capacity (IVC) was assessed too. Plethysmographic lung volumes, such as Total Lung Capacity (TLC), Intra Thoracic Gas Volume (ITGV), Expiratory Reserve Volume (ERV) and Residual Volume (RV) were assessed. Inspiratory Capacity(IC)/TLC ratio [15] was calculated by dividing TLC − ITGV for TLC, and was used as a marker for hyperinflation (defined as an abnormal increase in the volume of air remaining in the lungs at the end of spontaneous expiration [16]). Diffusing capacity of the lung for carbon monoxide (DL_CO) was measured using the single-breath method. All obtained values are expressed as percentages of the predicted value, by comparison with age and sex-specific reference values [17].

Body composition

Body height was measured to the nearest 0.5 cm with a wall-mounted stadiometer. Body weight was assessed to the nearest 0.1 kg using a weighing scale, and BMI was calculated as weight/height². A total body scan was performed by DEXA using a Lunar Prodigy® system (GE Healthcare - enCORE v14, Madison, WI, USA). Subjects lay supine on the DEXA table with arms adequately separated from the trunk and were instructed to remain still throughout the scanning procedure. After analysis of the whole body scan, a quadrilateral box was manually drawn around the L1–L4 region of interest (abdomen) bounded inferiorly by the horizontal line identifying L4/L5 vertebral space and superiorly by the horizontal line identifying the T12/L1 vertebral space. Scans were displayed with an adjustment of the gray scale, so that all of the soft tissue in the designated area was included. From the total body scan, body composition was assessed: bone mineral content (BMC), lean mass (LM), FFM (LM plus BMC). FFMI was calculated by dividing FFM by height², and FM by subtracting FFM from total weight. The base of the android region (Fig. 1a - shown in blue) sits immediately above the pelvis and is in height equal to 20 % of the distance from the pelvis to the chin. The android and gynoid (Fig. 1a - shown in red) regions are separated by a distance equal to 1.5 times the height of the android region, while the height of the gynoid region is double that of the android one. VAT was estimated within the android region [12] (Fig. 1b number 1).

Subjects characteristics

Patients were characterized by moderate to very severe airflow obstruction (GOLD I/II, 44.9 %; GOLD III, 38.1 %; GOLD IV, 17.0 %). 154 (52.4 %) were men (median age 67 years, IQR 13 years; median BMI 26.9 kg/m², IQR 8.6 kg/m²) and 140 (47.6 %) were women (median age 61 years, IQR 13 years; median BMI 25.3 kg/m², IQR 8,1 kg/m²). The median number of pack/ years smoked was 40 (IQR 24), 30.3 % of patients were current smokers and 63.6 % of patients were former smokers. 25.5 % of patients used long-term oxygen therapy.

VAT across GOLD stages

The amount of VAT separated for men and women is shown in Fig. 2a. In each GOLD subgroup VAT was greater in men than in women (in all GOLD stages: p < 0.001). Furthermore, patients with more severe airflow obstruction had significantly lower values of VAT, both in men (GOLD I/II: 2501.1 g ± 1141.7 g; GOLD III: 2246.4 g ± 1488.5 g; GOLD IV: 1697.2 g ± 1350.1 g, p =0.031) and women (GOLD I/II: 1259.7 g ± 938.0 g; GOLD III: 1160.0 ± 929.8 g, GOLD IV: 653.0 ± 442.8 g, p =0.015). Figure 2B shows the amount of VAT in the three different GOLD subgroups after stratification for BMI: as it could be expected, patients with a BMI ≥ 25 kg/m² had consistently higher values of VAT, compared with those with a BMI < 25 kg/m² (p < 0.001), both in men and women. After stratification for gender and BMI, differences in VAT across the three GOLD subgroups disappeared (Fig. 2b).

Relationships of VAT with other indices of total and central adiposity

The amount of VAT was significantly and strongly correlated with BMI, waist circumference, FM, FMI, android fat mass and gynoid fat mass (Fig. 3 a-f). The strength of correlations was higher in male than in female patients. Android fat mass was the most strongly correlated with VAT (r² = 0.919 in men; r² = 0.857 in women), followed by waist circumference (r² = 0.831 in men; r² = 0.757 in women), FM (r² = 0.858 in men; r² = 0.694 in women), FMI (r² = 0.894 in men; r² = 0.653 in women), BMI (r² = 0.826 in men; r² = 0.641 in women) and gynoid fat (r² = 0.746 in men; r² = 0.467 in women).

Impact of VAT on lung volumes

The same trend was observed in both genders, although it was less pronounced in female than in male patients (data not shown). DL_CO was significantly increased in patients with greater VAT in all GOLD categories (p = 0.002 in GOLD I + II; p = 0.003 in GOLD III; p = 0.015 in GOLD IV). The same finding was true when considering male and female patients separately.

Determinants of hyperinflation

VAT in COPD

Furutate et al. [19] investigated VAT in patients with COPD using CT. They observed that VAT was higher in COPD compared to controls with comparable BMI, and that VAT tended to increase with COPD progression, although not all differences were statistically significant. In addition, the authors found that, in patients with severe emphysema, VAT was retained despite the absence of obesity. In the same study, BMI, waist circumference, FFMI, FMI and SAT were shown to decrease with increasing emphysematous severity. Besides the fact that we did not relate VAT to emphysema severity, the differences in ethnicity (Furutate included a Japanese sample, the present study population were all caucasians) between studies makes comparison difficult, but our results indicates that VAT is comparable across GOLD stages, after stratification for BMI.

Assessment of VAT in COPD

Several methods can be used for measuring VAT. The most commonly used imaging technique is abdominal CT, one of the current gold standards for quantifying VAT [20–24]. However, because of the radiation dose involved, the high costs and the heavily utilized clinical equipment requirement, CT is not optimal as screening method for VAT. MRI has also been used for VAT measurement [25, 26]. It avoids the radiation dose involved with CT measurements but still requires an expensive and time-consuming procedure as well as heavily utilized clinical equipment. DEXA is another suitable clinical reference method for the assessment of body composition. It measures the differential attenuation of two different energy level x-rays as they pass through the body and subdivides soft tissue into bone, lean and fat compartments. New software was recently proposed to quantify VAT in the android region by DEXA [12, 13], with several advantage relative to CT and MRI, such as lower costs, lower x-ray exposure and short-scanning time [27–30].

Correlates of VAT

To date, no data are available on relationships between VAT measured by DEXA and other indices of body composition in COPD patients. In the present study we find a strong and significant correlation of VAT with BMI, waist circumference, FM, FMI as well as android and gynoid fat mass, in both genders but more strongly in males (likely due to their higher amount of VAT compared with females). As reported by Kaul et al., DEXA is a valid tool for quantification of VAT and other adipose tissue regions in the clinical setting [12], but more studies are needed in COPD in order to validate this method against CT, which is one of the current gold standards for quantifying VAT.

Hyperinflation in COPD

To the best of the authors’ knowledge, this study is the first that carefully evaluated VAT by DEXA in relation to static lung volumes in a large group of patients with moderate to very severe COPD. O’Donnell et al. [31] examined the relationship between increasing BMI and plethysmographic lung volumes in a population with airflow obstruction. It was shown that lung hyperinflation decreased exponentially with increasing BMI. However, the potential impact of fat distribution on lung volumes was not investigated. In the present study, patients with greater VAT seemed to have lower ITGV, ERV, RV and higher IC/TLC and DL_CO without a difference in FEV₁, FEV₁/FVC or IVC in most GOLD stages (and most pronounced in early GOLD stage). This finding appears to confirm that the major effect of obesity is on lung volumes, with no direct effect on airway obstruction [32]. The results of the multiple stepwise regression analysis confirmed a predicted role of FEV₁, BMI, gender and age on lung hyperinflation but no independent role of VAT for any of the lung volume parameters. This finding is somewhat surprising, since VAT is located in the abdominal region below the respiratory compartment. Future studies have to seek out what parameters of obesity are related to lung hyperinflation. Probably SAT in the abdominal and thoracic region could be worth considering.

In health, increasing BMI is known to be associated with increasing diffusing capacity, possibly reflecting the increased pulmonary blood volume in obesity [33, 34]. In the current study, a relationship was confirmed between increasing BMI and DL_CO in COPD patients. Nevertheless, no significant predictive role of VAT on DL_CO was found.

Clinical implications

VAT is now recognized as a risk factor for metabolic syndrome, diabetes, and cardiovascular disease [35–41]. Furthermore, since cardiovascular disease is one of the main comorbidities in COPD, together with type II diabetes, skeletal muscle dysfunction and osteoporosis [3], VAT is receiving increasing attention in clinical assessment of COPD patients.

DEXA provides lots of useful information in clinical practice. Not only is it the gold standard method for evaluating bone mineral density and discriminate patients with and without osteoporosis, but it is also a useful method for the assessment of other clinically relevant body composition parameters, among which VAT.