INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: cystic fibrosis; pulmonary function; pseudouridine

Chronic bacterial infection of the respiratory tract causing lung destruction and loss of pulmonary function is the major complication of cystic fibrosis (CF), which accounts for much of the morbidity and over 90% of the associated mortality (1, 2). Failure to maintain normal body weight and composition is also associated with increased morbidity and predicts a shortened life span (3). With the continuing improvement in survival, loss of fat-free mass (FFM) and reduced bone mineralization with an increased risk of fracture are becoming major complications in adults (5). Such studies suggest that alterations in body composition are related to the severity of lung disease. The mechanisms underlying this link are unknown, although it has been suggested that the host inflammatory and metabolic responses to chronic pulmonary infection may have an impact on maintenance of body composition (17).

It is likely that changes in body composition are related to a negative energy balance resulting from an inadequate energy intake to meet increased energy needs. A 25-80% greater energy requirement was demonstrated in patients with moderate to severe lung disease and nutritional depletion compared with age, sex, and size-matched healthy non-CF subjects (18). Excessive energy expenditure in CF may result from the gene defect (19), the increased energy cost of breathing due to altered pulmonary mechanics (4, 5, 20), and a catabolic intermediary metabolism secondary to chronic pulmonary infection (17, 21, 22). In adult patients, a mean resting energy expenditure (REE) of 121% predicted was associated with raised circulating catecholamines, lipolysis, altered body composition suggesting loss of skeletal muscle, and an acute phase inflammatory response with increased circulating immunoreactive tumor necrosis factor- (TNF-) (21).

A potential link between lung disease and body composition is the sustained host inflammatory response to chronic pulmonary infection. Acutely, the inflammatory response is accompanied by a parallel catabolic response, which is usually a short-term host-protective event (23). It is likely that proinflammatory, procatabolic cytokines, such as interleukins (IL)-1 and -6, TNF-, and counterregulatory hormones mediate the catabolic response with mobilization of fat and skeletal muscle as alternative energy sources (23). In CF the acute phase response is virtually continuous, particularly in adults with long-standing infection, and shows only modest reduction after specific antibiotic treatment (17, 21, 24, 25). Supporting the potential for such mechanisms acting in CF is the parallel reduction in the inflammatory and catabolic responses, and the reduction in bone collagen-1 breakdown after antibiotic treatment in adults with chronic Pseudomonas aeruginosa infection of the lung (17, 22).

We hypothesized that altered body composition in adults with CF may be due in part to the effects of altered pulmonary mechanics and chronic inflammation on intermediary metabolism leading to enhanced protein and lipid catabolism. To test this, we studied adults with chronic pulmonary infection with P. aeruginosa during a period of clinical stability to determine the relationships between the severity of lung disease, body composition, the host inflammatory response, dietary intake, and evidence of cellular and connective tissue protein breakdown.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We studied 21 males and 19 females with proven CF (17) and chronic pulmonary colonization with P. aeruginosa, mean (95% CI) age 23.1 (21.4, 24.7) yr. Twenty-two age-matched healthy subjects were also studied (11 males), mean (95% CI) age 23.8 (22.5, 25.1) yr. All subjects gave written informed consent and the study was approved by the Local Research Ethics Committee.

The patients were studied when clinically stable (no change in symptoms or medication and no reduction in FEV₁ of more than 10% of patient's usual value in the month previous to the study). Patients with diabetes mellitus, liver cirrhosis, or cor pulmonale, or receiving oral corticosteroid therapy, were excluded. All patients received chronic treatment with inhaled corticosteroids, 12 by nebulizer and 28 by inhaler. All patients received short acting ₂-agonist treatment by inhaler (n = 18) as required or by nebulizer (n = 22) on a regular daily regimen. All were prescribed vitamin D supplements.

Body Composition

Body composition was determined by dual-energy X-ray absorptiometry (DXA, Hologic 2000). The fat mass, FFM, and FFM index (FFMI) were determined (17). The FFM was expressed as either greater than or less than the lower 5th percentile for the matched healthy subjects (17). The coefficient of variation for bone mineral density (BMD) determinations at all sites and the body composition measurements ranged from 0.6 to 1.2%.

Nutritional Intake

Nutritional intake was assessed by a prospective self-completed 3-d food diary (26).

Bone and Intermediary Metabolism

Blood samples were collected at 9 a.m. after a 12-h fast. Serum insulin-like growth factor (IGF)-1, cortisol, bone-specific alkaline phosphatase (BSAP), parathormone (PTH), and vitamin D (25-hydroxycholecalciferol) were measured (17). Serum insulin was measured by immunochemiluminometric assay (Moleculat Light Technology Research Ltd, Cardiff, UK) and nonesterified fatty acids (NEFA) (Wako NEFA-C kit, Alpha Labs, Hants, UK) and triglycerides (Vitros Chemistry, Rochester, NY) by enzymatic colorimetric methods. The serum TNF-, IL-6, and their soluble receptors were measured by enzyme-linked immunosorbent assay (ELISA) (17).

A second void morning urine sample was collected. Cross-linked N-telopeptides of type I collagen (NTx) were determined (17). In the healthy subjects, agreement was found between NTx measured in 24 h and in second void "spot" morning urine samples (kappa 0.63). Pseudouridine (PSU) was measured by high-performance liquid chromatography (HPLC) and reported as ratio to urinary creatinine (27).

Other Measurements

FEV₁ was measured by dry wedge spirometry. The patients with FEV₁ < 45% predicted had severe impairment, FEV₁ > 46 and < 65% had moderate impairment, and FEV₁ > 66% had mild impairment of lung function (28). The number of exacerbations of the pulmonary symptoms during the previous year was obtained from the patient's notes. Physical activity was measured by questionnaire (17).

Statistics

Data were expressed as arithmetic mean (95% CI). Nonnormally distributed data were log₁₀ transformed and reported as geometric means. Comparison between groups was assessed by Student's t test and one-way ANOVA with Tukey's test. The paired t test was used for repeated measurements. Multiple stepwise regression was used to assess the impact of various factors on dependent variables.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary Function and Body Composition

The mean FEV₁% predicted (95% CI) of the 40 patients was 61.7 (54.3, 69.1)%. Based on FEV₁, 14 patients had severe, 11 moderate, and 15 mildly impaired lung function. There was no difference in the mean ages or sex distribution between the patients with severe and mild lung function impairment, p = 0.20 and p = 0.28, respectively. Of the 40 patients, 24 (12 males) had a low FFM and 16 had a normal FFM (9 males). The FEV₁ and FFM were related, r = 0.45, p = 0.004, and those with a low FFM had a lower mean (95% CI) FEV₁% predicted 53.9 (45.0, 62.8) compared with the normal FFM patients, 72.4 (60.8, 83.9) (p < 0.01). Stepwise multiple regression with FFM as the dependent variable revealed a significant effect of FEV₁ on FFM, r = 0.23, p = 0.002, but no effect of energy intake or circulating inflammatory mediators. None of the female patients was amenorrheic or oligomenorrheic and the male patients all had clinical signs of normal sexual development.

Pseudouridine Excretion

Excretion of PSU was greater in patients, 25.1 (18.2, 33.9) (mean [95% CI]), than in healthy subjects, 14.0 (9.5, 20.4) mmol/ mmol creatinine, p = 0.019. For the whole group of patients PSU was inversely related to FEV₁, r = 0.53, p = 0.001. Those with moderate or severe impairment had greater PSU excretion than those with mild impairment or the healthy subjects (Figure 1). To control for differences in body composition, PSU excretion was related to the FFMI (FFM/height²). The difference between the severe and mild groups was maintained; mean (95% CI) PSU:FFMI 2.80 (1.86, 3.37) mmol/ mmol/kg/m² and 1.06 (90.62, 1.49), respectively, p < 0.01. Patients with a low FFM had greater PSU excretion than those with a normal FFM, mean (95% CI) 33.9 (21.9, 51.3) compared with 16.2 (10.7, 24.5) mmol/mmol, or the healthy subjects, both p < 0.05. In those with a normal FFM, the PSU was similar to the healthy subjects.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Pseudouridine (PSU) (mmol/mmol creatinine) and NTx (nmol/mmol creatinine) in patients with severe, moderate, and mild impairment of lung function and healthy subjects, respectively; *p < 0.01 and **p < 0.05.

Bone Metabolism

The excretion of NTx:creatinine was greater in patients than in healthy subjects, 70.7 (57.5, 87.1) compared with 36.3 (30.3, 42.3) nmol/mmol creatinine, p < 0.01. The ratio of NTx:creatinine to BSAP activity was greater in patients, p < 0.05 (Table 1). Excretion of NTx was related to the severity of lung disease, 95.5 (63.1, 144.5) nmol/mmol creatinine in those with severe impairment compared with 54.9 (40.2, 75.8) nmol/mmol creatinine in those with mild impairment, p < 0.01, and 69.2 (51.3, 93.3) nmol/mmol creatinine in patients with moderate severity, p = 0.09, compared with mild severity (Figure 1).

Excretion of NTx was greater in patients with a low FFM compared with those with a normal FFM, 85.1 (66.1, 107.2) and 53.7 (37.2, 77.6) nmol/mmol creatinine, respectively, p < 0.05. The NTx:creatinine ratio was related to the FFM, to control for body composition differences, and remained greater in patients, p < 0.01 (Table 1). The NTx:creatinine ratio was related to bone mineral content (BMC) to allow for differences in bone mass between subjects and was increased in those with a low FFM, 48.7 (40.7, 79.5) nmol/mmol/g, compared with those with a normal FFM, 30.5 (11.8, 49.3), p < 0.05, or the healthy subjects, 10.9 (3.2, 18.7), p < 0.05.

Relationships between Bone Connective Tissue Protein and Cellular Protein Breakdown

In the patients as a group, NTx and PSU excretion were related, r = 0.33, p = 0.04. Multiple stepwise regression with PSU as a dependent variable revealed that FEV₁ (r = 0.31, p = 0.03) had a greater impact than FFM (r = 0.29, p = 0.06) or IL-6 (r = 0.02, p > 0.05). With NTx as the dependent variable, FEV₁ (r = 0.45, p < 0.01) again had a greater impact than FFM (r = 0.28, p = 0.07) or IL-6 (r = 0.27, p = 0.09).

Other Aspects of Bone Metabolism

The mean (95% CI) circulating vitamin D concentration for the patients was 18.2 (15.4, 21.1) ng/ml and was within the healthy reference range. Mean serum PTH concentrations for the patients were within the healthy reference range (17); mean (95% CI) 21.0 (17.8, 24.1) pg/ml, although 18 were above the upper 95% CI for healthy subjects. The total BMD and at all sites was less in patients than in healthy subjects (p < 0.01, data not shown). The Z scores of all but one patient showed osteopenia. Total BMD and NTx:creatinine were related in the patients group, r = 0.33, p = 0.03.

Inflammatory Mediators

Circulating IL-6, IL-6 soluble receptor, TNF-, and its soluble receptors 1 and 2 were all greater in patients than in healthy subjects (Figure 2). Only circulating IL-6 was related to the FEV₁, r = 0.46, p = 0.04. Severely impaired lung function was associated with greater concentrations of IL-6 (p < 0.01) and IL-6 soluble receptor (p < 0.05) than those with mild severity disease (Table 2). IL-6 and TNF- soluble receptor 2 (sr 2) levels were greater in patients with a low FFM than in those with a normal FFM. IL-6 was 6.02 (4.16, 6.19) and 2.75 (1.62, 4.67) pg/ml and TNF- sr 2 was 3019.9 (2630.3, 3388.4) and 2511.8 (2187.7, 2884.0) pg/ml for the low and normal FFM groups, respectively. Both these inflammatory mediators were related directly to PSU excretion, IL-6, r = 0.33, p = 0.047 and TNF- soluble receptor 2, r = 0.35, p = 0.031. TNF- soluble receptor 1 was related to NTx excretion, r = 0.42, p < 0.01. All patient subgroups had greater levels of inflammatory mediators than the healthy subjects (Table 2).

View larger version (15K): [in this window] [in a new window]   Figure 2.   IL-6 (pg/ml), TNF- (pg/ml), and TNF- soluble receptors (sr) 1 and 2 (ng/ml) in clinically stable patients and healthy subjects. *p < 0.01 and **p < 0.05.

Exacerbation Rates

Severe or moderate impairment of lung function was associated with more exacerbations in the year before the study than in patients with mild impairment (Table 2). A low FFM was associated with 4.7 (3.2, 6.3) mean (95% CI) exacerbations of pulmonary symptoms in the year preceding the study compared with 2.7 (2.0, 3.4) in those with a normal FFM, p < 0.01. The number of exacerbations in the previous year was related to FEV₁ (r = 0.63, p < 0.001), FFMI (r = 0.46, p < 0.01), PSU (r = 0.36, p = 0.02), NTx (r = 0.35, p = 0.03), IL-6 (r = 0.33, p = 0.03), and TNF- soluble receptor 2 (r = 0.40, p = 0.01).

Nutritional Assessment

Twenty-one patients (10 male) had low energy intake compared with that recommended (120-130% estimated average requirements [EAR] for age and sex). The mean (range) energy intake was 2354.9 (686.0, 5010.1) kcal/d, and the proportion of energy derived from fat, mean (range), was 36.8 (21.3, 51.6)%; protein, 14.5 (8.0, 30.4)%; and carbohydrate, 47.9 (36.1, 61.9)%. The mean (95% CI) total energy intake was 2281.5 (1783.0, 2780.0) and 2411.6 (1887.9, 2935.3) kcal/d, for the low and normal FFM groups, respectively, p > 0.05. Twenty patients (50%) had a lower protein intake than their EAR.

Hormone and Lipid Levels

Circulating cortisol, IGF-1, and the cortisol:IGF-1 ratio were similar in patients and healthy subjects, and within healthy reference ranges. There was no relationship to the severity of lung disease. Fasting insulin and glucose and the insulin:glucose ratio were also within normal limits (Table 1). Circulating NEFA and triglyceride levels were significantly greater in patients compared with healthy subjects (Table 1). There was no difference in hormone or lipid levels between patients with a low or normal FFM.

Physical Activity

Physical activity was less in patients with severe impairment of lung function compared with mild severity impairment, p < 0.05, and was related to the FEV₁, r = 0.44, p < 0.05 (Table 2). Patients with a low FFM were less physically active, 32.3 (26.4, 38.0) METS, than those with a normal FFM, 40.2 (36.1, 44.2) METS, p < 0.01. Physical activity was inversely related to the NTx:BMC ratio, r = 0.38, p = 0.017, but not to PSU excretion.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In clinically stable adults with CF and chronic P. aeruginosa infection of the lungs, there was excess breakdown of both cellular and connective tissue protein, which was related to the degree of impaired lung function and the systemic inflammatory response. The severity of lung disease and the systemic inflammatory response were both related to a low FFM and reduced BMD, the likely consequences of protein breakdown. These relationships indicate a parallel acute phase inflammatory and catabolic host response, which is probably continuous in clinically stable patients. These findings support our hypothesis and extend earlier suggestions that chronic lung disease is a major factor in altered body composition and clinically relevant metabolic complications in adults with CF (13, 15, 17, 20).

Impairment of pulmonary function had a greater effect on both cellular and connective tissue protein breakdown than the systemic inflammatory state, in that FEV₁ was better related to the excretion of PSU and NTx than IL-6, TNF-, or their soluble receptors. This was supported by multivariate analysis, although this requires cautious interpretation. These variables were also associated with the FFM, but less strongly than to the severity of impaired pulmonary function. An interpretation of this is that loss of lung function is a causative factor in the loss of FFM. A potential mechanism is indicated by the inverse relationship of increased excretion of PSU and NTx to pulmonary function. The inflammatory and parallel catabolic response is a candidate mechanism linking chronic lung infection with the metabolic complications of CF. This possibility is supported by the relationship between IL-6 and severity of lung disease and between IL-6 and TNF- soluble receptor 2 and PSU excretion. Hence, there is an association between the severity of lung disease, circulating proinflammatory and potentially procatabolic cytokines, protein breakdown, and FFM. A similar relationship also occurred for bone connective tissue protein breakdown and the severity of lung disease. The net breakdown of the organic matrix of bone was indicated by the imbalance when compared with BSAP, a marker of bone formation. Combined, these findings are strong circumstantial evidence of a link between these factors and loss of normal body composition in adults with CF.

This interpretation can be questioned as we measured immunoreactive cytokine, without knowing biological potency. However, the presence of natural inhibitors makes determination of cytokine bioactivity in body fluids just as imprecise. The circulating IL-6 soluble receptor concentration was increased in the patient groups and in the context of the known actions of IL-6 and its systemic endocrine effects our speculation can be justified (29, 30). The TNF- concentration was similar in both normal and low FFM groups, but the potential for bioactivity was indicated by the differential relationship of soluble receptor 2 concentration between the low and normal FFM groups (31). Our earlier finding that TNF- concentrations accounted for 30% of the variance for the increased REE in CF suggests potential bioactivity. Demonstration of a proteolytic myopathy in the transgenic IL-6 mouse and TNF- activation of NF-B-mediated myosin heavy chain isoform switching in vitro cultured muscle supports a potential proteolytic role in disease states (32, 33). Observations in other inflammatory disorders with metabolic consequences also support a possible role of cytokine involvement in the catabolic state (34). This catabolic state probably partly underlies altered body composition rather than a simple protein and calorie lack secondary to a poor intake. A reduced intake of energy leads to physiological adaptations to conserve energy and body composition until extreme starvation occurs, rather than the highly catabolic state related to lung disease in our patients (37).

Loss of FFM impairs inspiratory muscle function adding to the severity of pulmonary compromise (4, 5), generating a vicious circle of pulmonary inflammation and progressive lung injury causing an increased work of breathing (5, 22). Both increased respiratory work and inflammation add to the energy needs of such patients, which with an inadequate nutritional intake lead to a negative energy balance and an adaptive consumption of other substrates such as protein. This may have been occurring in our patients, in whom the total daily energy intake and the proportion as protein were deficient in over 50%. Hence, they were likely to be in a negative energy balance causing alternative substrate use, such as protein, despite an inadequate protein intake (20, 38). This would lead to breakdown of protein-rich tissues such as skeletal muscle and bone connective tissue. Our findings of a link between the severity of lung disease and body composition are supported by earlier studies in which loss of FFM was associated with a raised REE. Approximately 50% of the excess REE was accounted for by an increased oxygen cost of breathing and 30% by systemic inflammation (20, 21). These relationships were emphasized by the parallel reduction of inflammation and catabolism, reduction of the increased REE, and an approximate 1.5-kg weight gain, despite an unchanged energy intake, after antibiotic treatment (22).

Loss of systemic skeletal muscle and reduced physical activity may also affect the BMD, because less natural stress on bone impairs maintenance of BMD (39). The interpretation of bone metabolism, based on the markers we measured, needs to be cautious, as other factors may have an impact, including low vitamin D levels and raised PTH, which were to some degree present in our patients (17). However, our findings are consistent with vertebral biopsy evidence of predominant bone breakdown in CF (40). Other factors contributing both to loss of skeletal muscle mass and BMD include possible malabsorption, corticosteroid therapy, reduced sex hormone levels, and hypoinsulinemia. These mechanisms probably act by an addition to an overall negative energy balance (22).

We studied clinically stable patients, but the importance of exacerbations was demonstrated by the relationship between both PSU and NTx, when clinically stable, and the number of exacerbations in the preceding year, which were also associated with more intense systemic inflammation and more severe lung disease. The implication is that patients with the severest impairment in lung function have a chronic background inflammatory-catabolic response, which is enhanced during exacerbations that occur more frequently in such patients. Hence, they may not recover lost protein mass between exacerbations and are in a vicious cycle of progressive decline. There was no imbalance between circulating IGF-1 and cortisol levels, although in a subgroup, an exacerbation increased NTx excretion was related to a procatabolic hormone imbalance (17). In addition, our patients were lipolytic, which confirms an earlier finding in CF and in catabolic patients with AIDS (21, 41). Lipolysis in this setting may indicate energy-wasting "futile cycling" of substrates between fat stores and the liver without energy generation. The proinflammatory, procatabolic cytokines TNF-, IL-1, and IL-6 stimulate hepatic lipogenesis and peripheral lipolysis (34), and may have a role in the impaired use of stored lipids, which increases the utilization of protein as an alternative energy substrate.

Limitations of This Study

This study could be criticized for only using indicators of protein breakdown rather than determining protein turnover. The latter is technically difficult and is impractical for large-scale studies or where repeated measures are needed and information is not obtained about the protein metabolism in specific tissues (42). We chose PSU and NTx to give insights into protein breakdown from cellular and bone connective tissue origins, and because loss of FFM and BMD are major complications in adults. Excretion of PSU indicates RNA breakdown, cell destruction, and protein catabolism. Its excretion was validated against the continuous infusion of L-methyl-2H³-leucine and was used to assess protein catabolism in disorders associated with weight loss and cachexia, such as human immunodeficiency virus (HIV) where it predicted disease progression (43, 44). Estimated energy and substrate intake by prospective recording over 3 d could also be criticized as less precise than weighing portions and may explain the lack of a relationship with FFM status. However, weighing portions can distort intake patterns (45). Despite limitations, the questionnaire results supported previous findings of poor nutritional intake in CF (18, 22).

The relationships between the severity of lung disease, systemic inflammation, loss of FFM, and increased PSU and NTx excretion indicate a role for chronic lung disease in the nonrespiratory complications of CF in adults and indicate potential underlying mechanisms. This has provided strong circumstantial evidence to support our hypothesis that lung disease in adults with CF is a major contributor to clinically relevant abnormalities of body composition. Although this study cannot define causative mechanisms, it indicates potential pathophysiological relationships, which could be investigated further in interventional studies.