Associations between BODE Index and Systemic Inflammatory Biomarkers in COPD

Abstract

Background: COPD is a multicomponent disease and systemic inflammation represents one of the possible mechanisms responsible for its systemic manifestations, including skeletal muscle weakness and cachexia. Fat-free mass index (FFMI) that reflects the skeletal muscle mass, has been shown to be associated with both dyspnoea and exercise capacity. We hypothesized that the multidimensional BODE index, that reflects the multicomponent nature of COPD, might be related to biomarkers of systemic inflammation. We further evaluated associations between FFMI and systemic inflammation. Methods: BODE index and FFMI were calculated in 222 stable COPD patients and 132 smokers or ex-smokers with normal lung function. Systemic inflammation was evaluated with the measurement of leptin, adiponectin, CRP, IL-6, and TNF-α in serum samples of COPD patients. Results: In patients with COPD, both BODE index and FFMI presented significant positive and negative associations respectively with leptin levels (R² 0.61 and 0.65, respectively), whereas FFMI presented an additional negative association with the levels of TNF-α (R² 0.38). No significant associations were observed in smokers or ex-smokers with normal lung function. Conclusions: Both BODE index and FFMI, are related to the circulating levels of leptin in patients with COPD, suggesting a possible role for leptin in the systemic component of COPD. The additional association of FFMI with TNF-α may further support a role of systemic inflammation in muscle wasting in COPD.

Keywords: :

• COPD
• Cytokines
• CRP
• Adipose tissue
• Nutrition
• FFMI.

INTRODUCTION

Chronic obstructive pulmonary disease (COPD) is characterized by airflow limitation that is not fully reversible and is associated with an abnormal inflammatory response of the lungs to noxious particles and gases (1). Traditionally, forced expiratory volume in 1 second (FEV₁) has been the main measure of COPD severity and still has a central role in guidelines. However, COPD is a multicomponent disease, and a more composite measure might provide further information beyond airway obstruction and limitation. BODE index represents the most studied multidimensional measure, which was originally designed to predict mortality in COPD (2).

Three of its components are considered either directly (Body mass index [BMI]) or indirectly (6-minute walking distance [6MWD] and Medical Research Council [MRC] dyspnoea scale) as possible determinants of muscle weakness, since they evaluate body mass (3) and exercise capacity (4). BMI still represents the easiest measurement for the evaluation of body composition and/or skeletal muscle strength (5). However, fat-free mass (FFM) index (FFMI), which is calculated as FFM/height square, reflects better the skeletal muscle mass (6), is related to both dyspnoea and exercise capacity (7) and is an independent predictor of survival in COPD patients (8, 9).

Systemic inflammation is considered a hallmark of COPD and one of the key mechanisms responsible for the increased rate of co-morbidities, including skeletal muscle weakness and cachexia (10). Adipose tissue is a highly active organ, and there is evidence that it secretes a large variety of proteins, including cytokines, chemokines, and hormone-like factors, including leptin and adiponectin (11). Leptin is a circulating hormone produced by adipose tissue that regulates several metabolic and inflammatory functions, both centrally and peripherally. Adiponectin has anti-inflammatory properties, by reducing inflammatory cytokines and inducing anti-inflammatory ones. Altered levels of leptin have been reported in stable COPD, as well as in COPD exacerbations (12, 13). Limited data are currently available on the role of adiponectin in COPD, with the exception of an increase in its levels in underweight COPD patients and a marginal difference between stable phase and exacerbation (14, 15).

We hypothesized that the multidimensional BODE index, that reflects better the multicomponent nature of COPD, would present significant associations with biomarkers of systemic inflammation, especially those related to adipose tissue. As a secondary outcome we evaluated whether FFMI could reflect the systemic inflammatory process of the disease.

METHODS

Patients

Unselected stable COPD patients (n = 222) were included from the outpatient clinics of 3 tertiary hospitals during a period of 3 years (2007–2009). The diagnosis and classification of COPD was based on GOLD guidelines (1). Inclusion criteria were proper treatment according to the stage of their disease and no self-reported asthma or atopy. Patients were excluded if they had a respiratory infection in the last 8 weeks, use of systemic corticosteroids or reversibility >12% plus 200 ml of airway obstruction after administration of a β₂-agonist. The control group comprised of 132 smokers/ex-smokers with normal lung function. These subjects were recruited during a pre-operative evaluation of their lung function status for a non-malignant upper abdominal operation, which does not have an impact on body composition and simultaneously does not represents a significant co-morbidity. All procedures were performed before the operation. The study was approved by the scientific committees of the 3 hospitals and all subjects provided informed consent.

Study variables

Charlson index

Baseline co-morbidities were quantified with the use of the Charlson's Co-morbidity Index Score (CCIS) (16). This score is the most extensively studied method of co-morbidity measurement (17). In the present study, we did not include as CCIS variables COPD, dementia and hemiplegia, since COPD patients are the main group of our study and patients suffering from dementia and/or hemiplegia are not eligible for the evaluation of significant components of BODE index.

Body composition assessment

BMI was calculated as weight/height² (Kg/m²). FFM was measured by bioelectrical impedance analysis using a BIA 101 system analyzer (Akern, Florence, Italy) with an operating frequency of 50 KHz at 800 μA, as previously described (18). FFM was standardized for height and expressed as FFM/height² (FFMI) (19).

Pulmonary function tests

FEV₁ and Forced vital capacity (FVC) were measured using commercially available spirometers (Master Screen Body, Viasys Healthcare, Jaeger, Hoechberg, Germany), according to the American Thoracic Society guidelines (20).

Dyspnoea and exercise capacity

Chronic dyspnoea was assessed using the modified MRC scale (21). Exercise capacity was assessed with the 6MWD according to the American Thoracic Society guidelines in a walking cross of 50 m (22). Patients on long term oxygen treatment (LTOT) performed the test with supplementary oxygen. All tests were supervised by an experienced pneumonologist. Oxygen saturation and pulse rate were recorded using a finger-adapted pulse oximeter.

BODE index

BODE index was calculated as previously described (2). It consists of four components with total score ranges from 0–10. In detail, the BODE index is calculated as follows: BMI: ≥21 Kg/m² = 0, <21 Kg/m² = 1. MRC scale: 0–1 = 0, 2 = 1, 3 = 2, 4 = 3. 6MWD: ≥350 m = 0, 250–349 m = 1, 150–249 m = 2, ≤149 m = 3. FEV₁% pred: ≥ 65 = 0, 50–64 = 1, 36–49 = 2, ≤ 35 = 3.

Inflammatory biomarkers

Blood samples were drawn and were centrifuged at 1500 g for 15 min at 4^°C and stored at −80^°C. C reactive protein (CRP) was measured using highly sensitive nephelometry (Da de Herring 035041, Marburg, Germany), with normal values <0.3 mg/dl. Plasma leptin and adiponectin and serum tumor necrosis alpha (TNF-α) and interleukin-6 (IL-6) were measured by enzyme-linked immunosorbent assays (R&D systems, Abington, UK). Limits of detection were 7.8 pg/ml, 0.246 ng/ml, 0.12 pg/ml and 0.039 pg/ml, respectively. For leptin and adiponectin, samples were further diluted 1:100, according to manufacturer's guidelines, so that the minimum detectable levels were 0.78 ng/ml and 0.025 μg/ml, respectively.

Statistical analysis

Data are expressed as mean ± SD or as median (interquartile ranges) for normally distributed and skewed data, respectively. Comparisons between COPD patients and controls were performed with Mann Whitney U-test and unpaired Student's t-tests, for skewed and normally distributed variables, respectively, whereas comparisons of proportions were performed using chi-square tests. Linear regression analyses were performed in both groups separately in order to evaluate possible associations between BODE index and inflammatory biomarkers.

Regression analyses were performed after adjustment for age, gender, smoking habit (on the basis of current or ex-smoking status and in pack-years), CCIS, hospitalizations and exacerbations not requiring hospitalization in the previous year and treatment regimens including LTOT. Adjustments for treatment regimens were performed only in COPD patients. Similar regression analyses were also performed between FFMI and inflammatory biomarkers. Data were interpreted as standardized coefficients with 95% confidence intervals (CI). Values that were not normally distributed were log-transformed in order to obtain normal distribution for the regression analysis. Significant associations were introduced in a stepwise model in order to define the most significant ones. P-values <0.05 were considered statistically significant. Analysis was performed with Graph Pad Prism (La Jolla CA, USA) and SPSS 15 (SPSS, Chicago, IL, USA).

RESULTS

The demographic characteristics of the study participants are presented in Table 1. COPD patients and control subjects did not differ in age, gender, smoking habit and BMI, but COPD patients presented lower FFMI, spirometry values and exercise capacity, whereas they presented more dyspnoea, more co-morbidities and a higher number of self-reported hospitalizations in the previous year. BODE index was significantly higher in COPD patients compared to smokers or ex-smokers with normal lung function, while FFMI was significantly lower (Table 1).

Table 1  Demographic characteristics of study participants

Variables	COPD Patients n = 222	Smokers/ex smokers with normal lung function n = 132	p-values
Age	63 (58–71)	60 (53–65)	0.450
Gender (female/male)	53/169	35/97	0.723
Smokers (current/ex)	135/87	82/50	0.645
Duration of COPD (years)	9 (5–14)	N/A	N/A
Smoking (pack-years)	56 (50–65)	51 (43–62)	0.250
BMI Kg/m^2	27.5 ± 4.0	28.0 ± 5.0	0.630
FFMI Kg/m^2	17.2 (15.6–20.0)	20.6 (19.0–22.5)	<0.001
FMI Kg/m^2	9.1 (6.9–12.1)	7.7 (5.9–9.6)	0.100
Post B/D FEV1 (% pred.)	50 (36–62)	94 (86–98)	<0.001
Post B/D FEV1/FVC (%)	56 (48–65)	82 (78–91)	<0.001
Dyspnoea MRC	2 (2–3)	0 (0–1)	<0.001
6MWD in meters	280 (226–349)	490 (450–633)	<0.001
Charlson index	3 (1–5)	1 (0–3)	0.002
Number of hospitalizations in the previous year	0.6±0.2	0.2±0.3	0.001
Non-hospitalized exacerbations in the previous year	1.1±0.3	N/A
Treatment regimens
ICS	85	—
LABA	149	—
Tiotropium	178	—
SABA	167	—
Theophylline	12	—
LTOT	15	—
BODE index	4.0 (2.0–6.0)	0 (0–1)	<0.001

Normally distributed data are presented as mean± SD, whereas skewed data are presented as median (interquartile ranges). Treatment regimens are expressed as absolute values. Abbreviations: BMI: Body mass index; B/D: bronchodilation; FEV₁: Forced expiratory volume in 1 second; FVC: Forced vital capacity; FFMI: Fat-free mass index; FMI: Fat mass index; MRC: medical research council; N/A: Not available; ICS: inhaled corticosteroids; LABA: long acting β₂-agonists; LTOT: long-term oxygen therapy; SABA: short acting β₂-agonists; Significant p-values are all in favor of COPD patients.

The levels of systemic inflammatory biomarkers are presented in Table 2. Patients with COPD had significantly higher values of TNF-α, IL-6, CRP and leptin compared to control group, whereas the levels of adiponectin did not differ significantly between the two groups (Table 2).

Table 2  Inflammatory variables in the study participants

Variables	COPD Patients n = 222	Smokers/ex smokers with normal lung function n = 132	p-values
TNF-a (pg/ml)	2.9 (2.0–4.5)	1.5 (1.0–1.6)	<0.001
IL-6 (pg/ml)	8.2 (6.5–10.9)	6.5 (5.0–7.0)	0.003
CRP (mg/dl)	0.60 (0.30–0.90)	0.20 (0.14–0.31)	0.002
Leptin (ng/ml)	9.5 (7.8–12.3)	8.0 (7.0–9.5)	0.001
Adiponectin (μg/ml)	4.5 (2.0–6.0)	5.0 (2.7–5.5)	0.145

Data are presented as median (interquartile ranges). Abbreviations: CRP: C reactive protein; IL-6: Interleukin-6; TNF-α: Tumor necrosis alpha; Significant p-values are all in favor of COPD patients.

Associations of BODE index and FFMI with systemic inflammatory variables

In the linear regression analysis for COPD patients, after the aforementioned adjustments, BODE index presented a significant positive association with leptin. FFMI presented a negative association with leptin and an additional negative association with TNF-α (Table 3). The forward stepwise linear regression model showed that leptin was the most significant predictor of low FFMI (R² = 0.71, p<0.001). No significant associations were observed in control group for either indexes. Correlations between both BODE index and FFMI with leptin values in patients with COPD, are provided in Figures 1A and 1B.

Using linear regression analysis after the aforementioned proper adjustments, we evaluated associations of the four components of the BODE index separately with leptin values (Table 4). A significant positive association was found between MRC scale dyspnea level and leptin values, while a negative association was found between 6MWD and leptin values in COPD patients. Using a forward stepwise linear regression analysis, the most significant association was that of MRC scale [R² = 0.49]

Table 3  Associations between BODE and FFMI indexes and inflammatory biomarkers in patients with COPD

	BODE Index			FFMI
Biomarkers	β (95% CI)	Adjusted R^2*	p-value	β (95% CI)	Adjusted R^2*	p-value
CRP	−0.05 (−0.07, 0.10)	0.003	0.175	−0.008 (−0.70, 1.06)	0.009	0.360
TNF-α	0.04 (−0.06, 0.70)	0.002	0.100	−0.29 (−0.32, 0.42)	0.380	0.002
IL-6	0.08 (0.01, 0.13)	0.030	0.085	0.08 (−0.01, 0.14)	0.009	0.12
Leptin	0.70 (0.50, 0.88)	0.610	<0.001	−0.70 (−0.89, 0.63)	0.650	<0.001
Adiponectin	−0.02 (−0.04, 0.05)	0.004	0.485	−0.04 (−0.07, 0.07)	0.005	0.510

*After proper adjustments for age, gender, smoking habit (current and ex-smoking and pack-years), Charlson index, hospitalizations, non-hospitalized exacerbations and treatment regimens. Abbreviations: CI: Confidence intervals; CRP: C reactive protein; IL-6: Interleukin -6; TNF-α: Tumor necrosis alpha.

DISCUSSION

In this cross-sectional study we have shown a significant association of BODE index with plasma levels of leptin in a well-characterized cohort of COPD patients. Additionally, FFMI presented a negative association with leptin and an additional, yet weaker, association with TNF-α. All the above observations were not observed in a control group of smokers and ex-smokers with normal lung funct- ion.

Figure 1A Correlation between BODE index and leptin values in patients with chronic obstructive pulmonary disease (COPD), using Pearson's correlation coefficient [r² = 0.66, p < 0.001].

The only multidimensional scoring system that has gained broad acceptance until today is the BODE index, which was initially developed as a prognostic marker for COPD patients, in an attempt to integrate not only the respiratory but also the systemic manifestations of COPD in a single grading system (2). Importantly, alterations in body composition can occur in COPD in the absence of clinically important weight loss, the most important being loss of FFM (23). FFM reflects the metabolically active organs, skeletal muscles being the largest of them (24).

In the present study, the BODE index presented a significant association with leptin levels. A similar significant association was also observed between leptin and FFMI. Leptin plays an important role in regulating energy balance, but its role in muscle wasting and cachexia is controversial (25, 26). The most plausible explanation for the association of leptin with BODE index is its implication in the systemic inflammatory process in COPD, and in particular its association with body composition indices and exercise capacity.

There is evidence that leptin is up-regulated during periods of catabolism in order to control energy expenditure (27), a situation that is present in COPD patients, and especially in those with significant cachexia and reduction of muscle mass. Another important observation is the possibility that increased systemic inflammation directly affects or regulates leptin metabolism. This link between systemic inflammation and leptin has led to the hypothesis that leptin is regulated by inflammatory cytokines, which in turn induce anorexia, energy imbalance, and loss of body mass, and in particular muscle mass (13, 25).

The latter probably leads to low exercise capacity and may partially explain the associations between leptin and both BODE index and FFMI. The above speculation is partially supported by the significant associations between MRC scale and 6 MWD with leptin values in patients with COPD, indicating that exercise capacity - which is expressed by the above two components - is the major predictor of the increased leptin values.

Figure 1B Correlation between FFMI values and leptin values in patients with chronic obstructive pulmonary disease (COPD), using Pearson's correlation coefficient [r² = −0.32, p < 0.001].

In the present study we have not been able to show any associations between adiponectin and BODE index or FFMI, despite the fact that a previous study suggested that underweight patients with COPD presented increased levels of adiponectin (14). This difference may be attributed to the higher number of patients included in our study, to the different methodology used for the estimation of associations (regression analysis and not a dichotomous procedure) and to the fact that in the present study we have taken into account several possible confounding factors in the regression analysis that had not been taken into account in the previous study (14).

Neither BODE or FFMI presented any significant associations with IL-6 in the present study, despite previous data suggesting that IL-6 may be associated with muscle weakness in an elderly population (28). The absence of associations between IL-6 and both indexes might be related to the fact that age was addressed as a confounding factor in our study. Increased systemic levels of TNF-α have been proposed as a possible mechanism of muscle atrophy and weakness (29). The positive association between FFMI and TNF-α may further support the fact that FFMI is more accurate in predicting muscle weakness compared to BMI.

Table 4  Associations between BODE components and leptin values in patients with COPD

	Leptin
BODE components	β (95% CI)	Adjusted R^2*	p-value
FEV1% predicted	−0,18 [−0.002,−0.01]	0.09	0.08
MRC scale [0–5]	0.31 [−0.008,0.04]	0.48	<0.001
6MWD in meters	−0.37 [−0.04,0.08]	0.43	<0.001
BMI Kg/m^2	0.03 [1.9.15]	0.006	0.452

*After proper adjustments for age, gender, smoking habit (current and ex-smoking and pack-years), Charlson index, hospitalizations, non-hospitalized exacerbations and treatment regimens. Abbreviations: CI: Confidence intervals; BMI: Body mass index; FEV₁: Forced expiratory volume in 1 second; MRC: medical research council; 6MWD: 6-minute walking distance.

Finally, the functional role of CRP is uncertain and controversial. The absence of any association with BODE index is quite surprising, since CRP was associated with both BMI and exercise capacity in a previous study (30). However, despite the fact that some adjustments were performed in that study, many confounding factors like co-morbidities and treatment regimens were not addressed. The non-significant association between CRP and FFMI was somehow expected, since many studies support a positive rather than a negative association between FFMI and CRP (31).

In our study several confounding factors were addressed in order to eliminate possible effects on both the systemic inflammation and on the components of BODE index and FFMI. We consider two of them as crucial. The first is the evaluation of co-morbidities. The presence of co-morbidities may up-regulate the formation of systemic inflammatory biomarkers; moreover, several co-morbidities may be involved in body composition through their metabolic effects (10).

The second is the evaluation of treatment regimens, and in particular inhaled corticosteroids, since they may affect the systemic inflammatory process (32), and may further lead to possible body mass alterations (33). Finally, the associations of FFMI with biomarkers reflecting the systemic inflammatory process may support the need for further longitudinal studies that will evaluate prospectively the role of a multidimensional index incorporating FFMI instead of BMI in the outcome of COPD patients. It could be argued that treatment with oral steroids during exacerbations could affect the interpretation of our results. However, considering that exacerbation frequency in our population was quite low, we do not believe that treatment with a low dose of oral steroids [30 mg] for a short period of time [up to 7 days] could have any effect on the final results.

A possible limitation of this study is the fact that the absence of associations in the control group might be attributed to the small number of control subjects, compared to the COPD patients.

In conclusion, we have shown that the multidimensional BODE index as well as FFMI are related to the circulating levels of leptin in patients with COPD. The additional association of FFMI with TNF-α further supports the closer association of FFMI with muscle wasting and cachexia in COPD and suggests that FFMI may represent a valuable parameter in the evaluation of the systemic inflammatory process.

DECLARATION OF INTEREST

All the authors declare that they have no conflict of interest related to the content of this manuscript. All authors are responsible for the content and writing of this paper.