Autonomic dysfunction

Study objective: To verify whether autonomic neuropathy (AN) complicating type I, insulin-dependent diabetes mellitus affected neuroadrenergic bronchopulmonary innervation.

Patients: Twenty nonsmoking diabetic patients without respiratory diseases were studied: 11 patients with AN (group AN) and 9 patients without AN (control; group C) diagnosed by standardized criteria.

Design: Patients underwent respiratory function tests and ventilatory scintigraphies with [sup.123]I-metaiodobenzylguanidine (MIBG) and with [sup.99m]Tc-diethylenetriaminepenta-acetic acid (DTPA) to assess both bronchopulmonary neuroadrenergic innervation and also permeability of the alveolar-capillary barrier to water-soluble tracers. Rates of pulmonary clearance of the two tracers were computed, and correlates were identified by nonparametric statistics.

Setting: University hospital.

Results: The AN and C groups had normal respiratory function test results and comparable duration of diabetes and quality of metabolic control. [sup.99m]Tc-DTPA clearance did not distinguish the groups, [sup.123]I-MIBG clearance was faster in the AN group than in the C group (mean [+ or -] SD half-time of the radiotracer time-activity curve [[T.sub.1/2], 116.1 [+ or -] 22.8 min in the AN group vs 139.5 [+ or -] 18.3 min in the C group, p = 0.022), which is consistent with neuroadrenergic denervation in the AN group, [sup.123]I-MIBG clearance was independent from [sup.99m]Tc-DTPA clearance. Faster [sup.123]I-MIBG clearance was significantly associated with worse performance in three of the four autonomic tests.

Conclusions: Neuroadrenergic bronchopulmonary denervation may occur in diabetic patients with AN despite normal clinical and respiratory function findings. Further research is needed to identify clinical and prognostic implications of these findings.

Key words: autonomic neuropathy; diabetes mellitus; neuroadrenergic bronchopulmonary innervation; pulmonary complications

Abbreviations: AN = autonomic neuropathy; DLCO = diffusion capacity of the lung for carbon monoxide; DTPA = diethylenetriaminepenta-acetic acid; GHb = glycosylated hemoglobin; MIBG = metaiodobenzylguanidine; ROI = region of interest; [T.sub.1/2] = half-time of the radiotracer time-activity curve

**********

Diabetes mellitus can affect both the structure and function of kidney, retina, and peripheral nerves. By impairing autonomic innervation and/or vascular supply, it can also compromise the function of the GI tract, heart, bladder, and reproductive organs. The lung is commonly considered to be spared from clinically significant diabetic complications. (1) Thickening of alveolar and capillary basal laminae, however, have occasionally been reported in patients with type I, insulin-dependent diabetes mellitus. (2,3) Respiratory function has been found to be normal by some authors, but not by others reporting impaired diffusion capacity of the lung for carbon monoxide (DLCO), and reduced pulmonary elastic recoil and lung volumes. (4-7) A depressed bronchoconstrictor response to both cholinergic stimuli and eucapnic hyperventilation with cold air has been observed and ascribed to reduced resting vagal activity. (8,9) Finally, a lower-than-normal threshold for dyspnea has been observed in diabetic patients during both hypoxic and hypercapnic rebreathing tests. (10,11) These findings are consistent with some kind of dysfunction of neuroautonomic mechanisms that regulate bronchial tone and control of breathing.

The present study was designed to comparatively assess the lung clearance of aerosolized [sup.123]I-metaiodobenzylguanidine (MIBG) in patients with type I diabetes mellitus with or without autonomic neuropathy (AN). Our basic hypothesis was that patients with AN are characterized by accelerated [sup.123]I-MIBG clearance, which is a sign of neuroadrenergic denervation. We used a method previously shown to be highly sensitive to changes in bronchial adrenergic tone. (12) We also performed a pulmonary ventilatory scintigraphy with [sup.99m]Tc-diethylenetriaminepentaacetic acid (DTPA) to verify whether the status of the alveolar-capillary permeability could explain the pattern of [sup.123]I-MIBG clearance.

MATERIALS AND METHODS

Subjects

Twenty patients with type I diabetes mellitus (9 male, 11 female; mean [+ or -] SD age, 44.30 [+ or -] 9.86 years) diagnosed according to standardized criteria were studied. (13) AN was diagnosed in 11 patients (AN group), and 9 patients did not have AN and were considered control patients (C group). AN was diagnosed according to a standardized procedure including four cardiovascular autonomic tests (Table 1). Normal, borderline, and abnormal responses to each maneuver of autonomic stimulation are reported in Table 1. Patients having at least one abnormal response were considered to have AN. According to Ewing and Clarke, (14) an autonomic cardiovascular score was computed by summing scores attributed to the four individual tests as follows: 0 = normal, 1 = borderline, and 2 = abnormal. Thus, the autonomic cardiovascular score can range from 0 to 8. (14) It must be noted that an autonomic score < 5 and > 1 is consistent with either a normal condition or the AN status, respectively. Indeed, a control patient can score up to 4 in the event of a borderline response to all four tests, whereas an AN patient can score only 2, corresponding to an abnormal performance in only one autonomic test. Only the analysis of the response to individual autonomic tests permits a diagnosis of AN.

The pulmonary clearance of IV-injected MIBG has been shown to be inversely related to age. (15) Although clearance of inhaled MIBG likely depends on different mechanisms, as will be discussed later, we decided to match for age the AN group and the C group in order to limit any potential source of confounding.

Methods

The investigation was conducted according to the principles expressed in the Declaration of Helsinki, and the patients gave informed consent to all procedures. The protocol was approved by the local ethical committee. Patients underwent a medical examination, fundoscopy and fluorescein angiography, ECG, chest radiography, and laboratory evaluation, including urinalysis, 12-parameter computerized serum multiple analysis, full blood count, glycosylated hemoglobin (GHb), C-reactive protein by a sensitive double-antibody sandwich enzyme-linked immunosorbent assay, and erythrocyte sedimentation rate. Criteria of exclusion from the study were the following: previous or current history of smoking; a standardized diagnosis of atopy (eg, a positive response to at least 1 of 12 prick tests of a standardized battery); either history, physical, radiologic, or echocardiographic findings consistent with heart or lung disease; liver disease; body mass index < 20 kg/[m.sup.2] or > 27 kg/[m.sup.2]; serum creatinine level > 1.2 mg/dL, anemia (hemoglobin < 13.5 g/dL in male patients and < 12 g/dL in female patients); abnormal serum level of C-reactive protein; erythrocyte sedimentation rate > 20 mm at the first hour; history of transient respiratory infection (cold included) in the 3 months prior to study; and use of drugs influencing the neuroadrenergic or cholinergic tone. All patients had to be in stable metabolic condition, as reflected by absence of glycosuria and normal glycemic 6-point profile performed in the week prior to the study. Patients with AN and control patients were matched according to age, occupational role, and lifestyle in order to have groups with comparable daily physical activity. Selecting criteria were aimed to prevent the confounding effect of metabolic decompensation, inflammatory diseases, abnormal nutritional status, and other conditions, such as the coexistence of physically active and sedentary patients, known or supposed to influence respiratory performance and/or the autonomous nervous system.

Respiratory Function Tests

Lung Volume and Flow: Lung volumes and flows were measured by computerized system (Med-Graphics 1070; Medical Graphics Corporation; St. Paul, MN). Spirometric performance had to meet American Thoracic Society criteria of acceptability and reproducibility of curves. (16) All values were expressed as percentage of the reference normal population. (17)

DLCO: DLCO was measured according to procedure recommended by the American Thoracic Society. (18) Two main parameters were obtained: DLCO, which reflects the overall transfer of carbon monoxide across the alveolar-capillary barrier; and coefficient of diffusion, a measure of diffusion per unit of alveolar volume. Reference normal values were derived from Cotes. (19)

Bronchial Reactivity: Bronchial reactivity to inhaled methacholine was performed according to the method proposed by Ryan et al, (20) using a dosimeter providing a calibrated output of 9.0 [micro]L per puff. The subjects inhaled an aerosol of diluent followed by double concentrations of methacholine from 0.031 to 16 mg/mL. The methacholine doses were administered at 5-min intervals, and FE[V.sub.1] was measured 30 to 90 s following each inhalation. A dose-response curve was drawn, and the slope of the curve was derived from the relationship between the FE[V.sub.1] fall, expressed in percentage from baseline, and the last dose inhaled. The methacholine slope curve was considered a more reliable index of bronchial reactivity in patients without signs of bronchial hyperreactivity, rather than the provocative concentration of methacholine that caused a 20% fall in FE[V.sub.1], which is probably more suitable for asthma studies.

Neuroadrenergic Bronchial Innervation

Pulmonary ventilatory scintigraphy with [sup.123]I-MIBG was performed to investigate the neuroadrenergic system of the lung. By sharing several pathways with noradrenaline in the adrenergic nerve terminal and being nonmetabolized, [sup.123]I-MIBG qualifies as a reliable marker of neuroadrenergic activity. Pharmacologically induced [beta]-blockade can significantly increase the lung clearance of [sup.123]I-MIBG, which is consistent with a larger fraction of [sup.123]I-MIBG becoming available for alveolar capillary transit as a consequence of reduced access to neuroadrenergic terminals. (12)

[sup.123]I-MIBG was nebulized by a pneumatic aerosol generator (Venticis II; Cis Diagnostici; Vercelli, Italy), and an aerosol having median aerodynamic diameter of 0.79 [+ or -] 0.06 [micro]m (geometric SD, 2.4 [+ or -] 0.08 [micro]m) was obtained. Such a diameter allows a high proportion of the aerosolized particles to reach the alveolar compartment. The cumulative dose of [sup.123]I-MIBG was 185 megabecquerels, and volume of the aerosol was 5 mL. Immediately after inhalation of the aerosol, the patient underwent a 20-min dynamic scintigraphic study in the supine position. A large-field gamma camera equipped with a low-energy, high-sensitivity collimator was placed in a posterior view below the scanning bed. Matrix size was 64 x 64, frame rate was 15 s per image, and total acquisition time was 20 min. After cumulating the first three images of the dynamic acquisition, two regions of interest (ROIs) corresponding to the hilar area and to extrahilar of the right lung (purely parenchymal) area, were defined. The ratio of the mean activity per pixel of the extrahilar to the hilar area was computed and represented the penetration index. Finally, a time-activity curve was obtained from the extrahilar ROI, and the [T.sub.1/2] of the curve was computed according to Rindercknecht et al (21) We excluded the left lung from computation in order to limit the confounding effect of the heart on the definition of hilar and extrahilar ROIs.

Alveolar Clearance of Water-Soluble Tracers

Inhaled [sup.99m]Tc-DTPA is commonly used to study the alveolar-capillary permeability. Its clearance is unaffected by changes in blood flow, and it is highly sensitive to several causes of parenchymal lung injury. (21) DLCO and clearance of [sup.99m]Tc-DTPA provide complementary information about the status of alveolar-capillary barrier. (22) A total of 740 megabecquerels of [sup.99m]Tc-DTPA was used, and the same procedure described for [sup.123]I-MIBG pulmonary scintigraphy was adopted for collecting scintigraphic data and computing the penetration index. A low-energy, general purpose collimator was used.

Two weeks elapsed between [sup.123]I-MIBG and [sup.99m]Tc-DTPA ventilatory scintigraphy. Both specialists in nuclear medicine performing the studies (A.G., M.L.C.) were blinded to the patients' group membership.

Statistical Analysis

Data were analyzed using the SX statistical package (Statistix, Version 4.0; Analytical Software; St Paul, MN). Given that the requisites of normal distribution and homogeneous variance could not be completely met, differences between groups were analyzed using the Mann-Whitney test. (23) The Spearman's [rho] test was used to assess the correlations between [sup.123]I-MIBG clearance and each of the following parameters: [sup.99m]Tc-DTPA clearance, GHb, and individual indexes of autonomic function. (24) Correlates of [sup.99m]Tc-DTPA clearance were also computed.

RESULTS

Table 2 shows demographic, clinical, and respiratory function data of the patients. Both groups had comparable and normal lung volumes and bronchial reactivity to methacholine. Three AN group patients but no C group patients had abnormal DLCO values. The quality of the metabolic control, as expressed by GHb, did not distinguish groups. Prevalence of proteinuria (> 3.5 mg/dL) and of diabetic retinopathy were significantly and tendentially higher in the AN group.

Table 3 shows the results of autonomic tests in both groups. Only two AN group patients achieved a score of 8 (an abnormal response to .all autonomic tests). The distribution of individual scores is consistent with a great variability in the severity of autonomic dysfunction characterizing the AN group. BP response to standing, which is the only test exploring the neuroadrenergic system, was abnormal in four patients, but was in the borderline range in the remaining seven patients. Autonomic scores in the C group ranged from 0 to 2, but were equal to 0 in five patients, equal to 1 in three patients, and equal to 2 in only one patient.

Table 4 summarizes results of radionuclide studies. [sup.123]I-MIBG clearance was significantly faster in the AN group than in the C group, whereas that of [sup.99m]Tc-DTPA was not significantly different between the two groups. However, a trend toward a slower clearance of [sup.99m]Tc-DTPA was observed in the AN group. It must be noted that when the AN group and the C group were pooled, [sup.123]I-MIBG and [sup.99m]Tc-DTPA had comparable penetration index (82.7 [+ or -] 8.7% vs 84.6 [+ or -] 9.6%, respectively; not significant). This finding is consistent with a comparable amount of both tracers reaching the lung parenchyma. Thus, the observed clearance of [sup.123]I-MIBG and [sup.99m]Tc-DTPA actually reflected the inherent clearance mechanisms and not the differences in distribution of the tracers.

Table 5 shows correlates of [sup.123]I-MIBG and of [sup.99m]Tc-DTPA clearance in pooled AN and C groups. The correlation between clearances of [sup.123]I-MIBG and [sup.99m]Tc-DTPA was not significant. Analogously, no correlation emerged between clearance of [sup.123]I-MIBG and GHb. Clearance of [sup.123]I-MIBG was significantly correlated with the total autonomic score, as well as with all but one autonomic tests. No significant correlation was found between clearance of [sup.99m]Tc-DTPA and autonomic tests. It must be noted that [sup.123]T-MIBG clearance was expressed by [T.sub.1/2], which is inversely proportional to the rate of clearance. Accordingly, the observed inverse correlation of [T.sub.1/2] with variables that were directly related to the severity of autonomic deficit, ie, the total autonomic score and BP response to standing, and the direct correlation with the remaining variables demonstrate that [sup.123]I-MIBG clearance was faster, which is consistent with neuroadrenergic denervation in patients with more severe autonomic deficit.

Figures 1, 2 show the four significant relationships between clearance of [sup.123]I-MIBG and selected indexes of autonomic dysfunction. The heart rate response to standing was the strongest correlate of [sup.123]I-MIBG clearance.

[FIGURE 1 OMITTED]

DISCUSSION

The present study demonstrates that loss of neuroadrenergic innervation of the lung might be a feature of AN complicating type I diabetes mellitus. The lack of correlation between alveolar permeability to water-soluble tracers and [sup.123]I-MIBG clearance guarantees that present findings really reflect neuroadrenergic denervation. The strong parallelism between the severity of AN, expressed by standardized indexes, and lung neuroadrenergic denervation further supports this conclusion. However, cholinergic bronchomotor tone did not distinguish diabetic patients with and without AN. The last finding is at variance with previous studies (25-27) showing depressed parasympathetic bronchomotor tone in patients with type I diabetes mellitus complicated by AN. However, Fonseca et al (25) studied diabetic patients with lower baseline FE[V.sub.1], expressed as percentage of predicted, than control normal subjects, whereas no information on the alveolar-capillary permeability was provided by either Bertherat et al (26) or Douglas et al. (27) This limits comparability of previous and present data. Our findings, obtained from subjects with normal respiratory function, suggest that neuroadrenergic precedes cholinergic dysfunction of the bronchomotor tone in the course of diabetic AN. A follow-up of our patients will test this hypothesis.

Diabetes mellitus is considered an independent risk factor for developing lower respiratory tract infections as well as a severity factor in pulmonary infections. (28) Longer hospital stay and larger percentage of Gram-negative bacilli in the sputum have been reported in diabetic patients with COPD and pneumonia. (29) Metabolic decompensation and selected immunologic problems, mainly a defective neutrophil function, likely qualify as a multiplier of the effects of acute respiratory infection. (30) However, it cannot be excluded that abnormal regulation of bronchomotor tone contributes to worsen prognosis of pneumonia in diabetic patients by impairing both defense reflexes of the airways and matching of ventilation to perfusion. (31) Furthermore, a neuroadrenergic dysfunction might favor the onset of or slow the recovery from acute respiratory infections, either by impairing the mucociliary clearance, which is physiologically stimulated by adrenergic stimuli, or by eliciting bronchospasm in predisposed subjects. (32) These theoretical considerations point to bronchopulmonary autonomic dysfunction as a possible severity factor of pneumonia in diabetic patients.

Respiratory function test results of our patients were normal. However, the observed trend toward slower clearance of [sup.99m]Tc-DTPA in patients with AN is consistent with the previously reported lower permeability of the alveolar-capillary barrier in diabetic patients with microvascular complications. (33) Furthermore, three AN group patients and no C group patients had abnormal DLCO values, although both groups had comparable and normal mean DLCO values. These findings suggest that neuroadrenergic denervation is likely the most evident, but not the only sign of impending respiratory dysfunction in diabetes mellitus complicated by AN.

An important conceptual issue deserves consideration: delayed lung clearance of IV-injected [sup.123]I-MIBG in diabetics has been considered to reflect some kind of endothelial dysfunction. (34,35) Theoretically, endothelial cells could contribute to the clearance of inhaled [sup.123]I-MIBG. Indeed, being an analog of the circulating biogenic amines (serotonin, prostaglandins, etc), IV-injected MIBG is captured by the endothelial cells of the pulmonary capillaries that metabolize these molecules in the lung. (34-37) However, inhaled MIBG is selectively deposited in the airways and primarily handled by the neuroadrenergic system of the airways; its uptake is reduced and its clearance accelerated by sympathetic blockade. (12) While a role of the capillary endothelium in the clearance of inhaled MIBG cannot be excluded, indirect evidence suggests otherwise. First, the lung clearance of inhaled "nonspecific" tracers is reduced in diabetic patients with complications as compared to those without. (33) We also report here such a pattern using [sup.99m]Tc-DTPA, although the difference between the groups did not reach statistical significance. Conversely, MIBG lung clearance was significantly accelerated in diabetic patients with neuropathic complications. Second, if endothelial uptake of MIBG had strongly interfered with our measurements, it would have delayed the MIBG lung clearance of neuropathic patients, as shown by Murashima et al (34) and Unlu and Inanir. (35) On the contrary, we observed that MIBG lung clearance in these patients was accelerated. Third, the parallelism between severity of autonomic dysfunction and clearance of inhaled MIBG as well as the previously demonstrated enhancement of MIBG clearance by [beta]-blockade both suggest that the neuroadrenergic system temporarily stores a consistent proportion of the inhaled tracer. (12) Thus, a faster clearance of inhaled MIBG likely reflects a decreased storage capacity (a defective neuroadrenergic system). These considerations suggest that mechanisms of clearance of MIBG vary depending on whether MIBG is inhaled or injected IV; therefore, previous and present studies are barely comparable. (34,35)

Caution is needed in interpreting the present findings for at least three reasons. First, the stringent selecting criteria probably allowed the exclusion of factors apt to confound the interpretation of results. However, patients with subclinical lung disease attributable to diabetes mellitus might also have been excluded. Theoretically, these patients might show more severe neuroadrenergic dysfunction as well as cholinergic abnormal tone and other respiratory function abnormalities. Second, we assessed neuroadrenergic innervation using a radioisotopic method, and cholinergic bronchomotor tone with a provocative pharmacologic test. This limits comparability of results inherent to the two arms of the neurovegetative system. It cannot be excluded that abnormal cholinergic denervation might have been detected by a dedicated tracer. However, at least as far as we know, such a tracer is unavailable. Third, the sample size was small, and a control group of normal subjects was lacking. However, MIBG clearance of diabetic patients without AN was fairly comparable to that previously measured in normal subjects. (12) Thus, the difference in MIBG clearance between diabetic patients with and without AN is unlikely to be due to chance.

These limitations do not substantially weaken the meaning of this study: neuroadrenergic dysfunction of the lung is likely to be a further effect of AN complicating type I diabetes mellitus. Future research should verify whether it worsens in parallel with progression of clinically evident extrapulmonary AN, as well as whether a break point in its natural history can be identified marking the onset of abnormalities in selected respiratory function tests. Finally, the hypothesis that neuroadrenergic dysfunction qualifies as a risk factor for defective mucociliary clearance or for resistance to bronchodilators seems worth testing. It should, furthermore, be verified whether our observation applies also to patients with type II diabetes mellitus, who represent the majority of diabetics in western countries and are highly exposed to the risk of AN. In conclusion, present data, although preliminary in nature, open a large spectrum of opportunities for further research into the relationship between functional status of lung innervation and respiratory diseases with potential implications for diabetics as well as for other categories of patients.

ACKNOWLEDGMENT: This article is dedicated to Dr. Patrizia Cotroneo.

REFERENCES

(1) Teplick R, Fishman AJ. Systemic infection, the sepsis syndrome, and the lungs. In: Fishman AP, ed. Pulmonary diseases and disorders. 3rd ed. New York, NY: McGraw-Hill, 1998:2071-2094

(2) Vracko RA. A comparison of the microvascular lesions in diabetes mellitus with those in normal aging. J Am Geriatr Soc 1982; 30:201-205

(3) Vogt BW, Schleicher ED, Wieland OH. Epsilon-amino-lysine bound glucose in human tissue obtained at autopsy. Diabetes 1982; 31:1123-1127

(4) Sandler M. Is the lung a "target organ" in diabetes mellitus? Arch Intern Med 1990; 150:1385-1388

(5) Ramirez LC, Dal Nogare A, Hsia C, et al. Relationship between diabetes control and pulmonary function in insulin-dependent diabetes mellitus. Am J Med 1991; 91:371-376

(6) Maccioni FJ, Colebatch HJ. Lung volume and distensibility in insulin-dependent diabetes mellitus. Am Rev Respir Dis 1991; 143:1253-1256

(7) Wanke T, Formanek D, Auinger M, et al. Inspiratory muscle performance and pulmonary function changes in insulin-dependent diabetes mellitus. Am Rev Respir Dis 1991; 143:97-100

(8) Pieron M, Scheen PM, Corhay JL, et al. Bronchial reactivity in diabetic patients. Rev Mal Respir 1997; 14:379-385

(9) Heaton RW, Guy RJ, Gray BJ, et al. Diminished bronchial reactivity to cold air in diabetic patients with autonomic neuropathy. BMJ 1984; 289:149-151

(10) Mancini M, Filippelli M, Seghieri G, et al. Respiratory muscle function and hypoxic ventilatory control in patients with type I diabetes. Chest 1999; 115:1553-1562

(11) Scano G, Filippelli M, Romagnoli I, et al. Hypoxic and hypercapnic breathlessness in patients with type I diabetes mellitus. Chest 2000; 117:960-967

(12) Giordano A, Calcagni ML, Rossi B, et al. Potential use of iodine-123 metaiodobenzylguanidine radioaerosol as a marker of pulmonary neuroadrenergic function. Eur J Nucl Med 1997; 24:52-58

(13) National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979; 28:1039-1057

(14) Ewing DJ, Clarke BF. Diagnosis and management of diabetic neuropathy. BMJ 1982; 285:916-918

(15) Kanzaki N, Soda R, Takahashi K, et al. Influence of age and gender on iodine-123 MIBG kinetics in normal lung. Ann Nucl Med 1998; 12:225-229

(16) American Thoracic Society. Standardization of spirometry: 1994 update. Am J Respir Crit Care Med 1995; 152: 1107-1136

(17) Knudson RJ, Lebowitz MD, Holberg CJ, et al. Changes in the normal expiratory flow-volume curve with grow and aging. Am Rev Respir Dis 1983; 127:734-753

(18) American Thoracic Society. Single-breath carbon monoxide diffusing capacity (transfer factor): recommendations for a standard technique; 1995 update. Am J Respir Crit Care Med 1995; 152:2185-2198

(19) Cotes JE. Lung function: assessment and applications in medicine. 4th ed. Oxford, UK: Blackwell Scientific Publishers, 1979

(20) Ryan G, Dolovich MB, Roberts RS, et al. Standardization of inhalation provocation tests: two techniques of aerosol generation and inhalation compared. Am Rev Respir Dis 1981; 123:195-199

(21) Rinderknecht J, Shapiro L, Krauthammer M, et al. Accelerated clearance of small solutes from the lungs in interstitial lung disease. Am Rev Respir Dis 1980; 121:105-117

(22) Pistelli R, Maini CL, Fuso L, et al. Pulmonary involvement in progressive systemic sclerosis: a multidisciplinary approach. Respiration 1987; 51:296-306

(23) Mann HB, Whitney DR. On a test of whether one of two variables is stochastically larger than the other. Ann Math Statist 1947; 18:50-60

(24) Olds EG. The 5% significance levels for sums of squares of rank differences and a correction. Ann Math Stat 1949; 20:117-118

(25) Santos e Fonseca CM, Manco JC, Gallo L Jr, et al. Cholinergic bronchomotor tone and airway caliber in insulin-dependent diabetes mellitus. Chest 1992; 101:1038-1043

(26) Bertherat J, Lubetzki J, Lockhart A, et al. Decreased bronchial response to methacholine in IDDM patients with autonomic neuropathy. Diabetes 1991; 40:1100-1106

(27) Douglas NJ, Campbell IW, Ewing DJ, et al. Reduced airway vagal tone in diabetic patients with autonomic neuropathy. Clin Sci 1981; 61:581-584

(28) Koziel H, Koziel MJ. Pulmonary complications of diabetes mellitus. Pneumonia Infect Dis Clin North Am 1995; 9:65-96

(29) Loukides S, Polyzogopoulos D. The effect of diabetes mellitus on the outcome of patients with chronic obstructive pulmonary disease exacerbated due to respiratory infections. Respiration 1996; 63:170-173

(30) Rayfield E, Ault M, Keusch G, et al. Infection and diabetes: the case for glucose control. Am J Med 1982; 72:439-450

(31) Hansen LA, Prakash UB, Colby TV. Pulmonary complications in diabetes mellitus. Mayo Clin Proc 1989; 64:791-799

(32) Sanderson MJ, Dirkensen ER. Mechanosensitive and [beta]-adrenergic control of the ciliary beat frequency of mammalian respiratory tract cells in culture. Am Rev Respir Dis 1989; 139:432-440

(33) Caner B, Ugur O, Bayraktar M, et al. Impaired lung epithelial permeability in diabetics detected by technetium-99m-DTPA aerosol scintigraphy. J Nucl Med 1994; 35:204-206

(34) Murashima S, Takeda K, Matsumura K, et al. Increased lung uptake of Iodine-123-MIBG in diabetics with sympathetic nervous dysfunction. J Nucl Med 1998; 39:334-38

(35) Unlu M, Inanir S. Prolonged lung retention of Iodine-123-MIBG in diabetic patients. J Nucl Med 1998; 39:116-118

(36) Block ER, Stalcup SA. Metabolic functions of the lung: of what clinical relevance? Chest 1982; 81:215-223

(37) Slosman DO, Polla BS, Donath A. [sup.123]I-MIBG pulmonary removal: a biochemical marker of minimal lung endothelial cell lesions. Eur J Nucl Med 1990; 16:633-637

* From the Departments of Internal Medicine (Drs. Antonelli Incalzi, Pitocco, and Ghirlanda), Respiratory Physiology (Drs. Fuso and Maiolo), and Nuclear Medicine (Drs. Giordano and Calcagni), Catholic University, Rome, Italy.

Manuscript received February 15, 2001; revision accepted July 18, 2001.

Correspondence to: Leonello Fuso, MD, Fisiopatologia Respiratoria, Universita Cattolica S. Cuore, Largo A. Gemelli 8, 00168 Roma, Italy; e-mail: leofuso@rm.unicatt.it

COPYRIGHT 2002 American College of Chest Physicians
COPYRIGHT 2002 Gale Group