Pulmonary Function in Young Insulin-dependent Diabetic Subjects

Pulmonary Function in Young Insulin-dependent Diabetic Subjects* Malcolm Sandler, M.R.C.P.;t Anthony E. Bunn, Ph.D.t. and Rogerl. Stewart, M.B., Ch.B., Ph.D.t.

To clarify the issue of pulmonary dysfunction in diabetes mellitus, lung mechanics and CO transfer were investigated in U young (mean age 19.5 :t 5 years) non-smoking, insulindependent diabetic patients and an equal number of matched healthy subjects. Mean closing capacity/total lung capacity (CC/TLC) was significantly greater in the diabetic than in the control group (31.4::1:6.8 vs 27.2::t:2.9 percent, p<0.01~ as was the mean value of the volume independent index of lung elasticity (exponent constant, Kst(L)) (0.148::1:0.045 vs 0.118::t:0.030, p<0.05~ 11ie transfer factor expressed per unit alveolar volume (TIJVA) was also signi6-

cantly lower in the diabetic than in the control group (5.25::1:0.68 vs 5.61±0.57 ml/min/mm Hg/L, p<0.05) and this could be ascribed to a lower pulmonary capillary blood volume. There was evidence of mildly abnormal lung mechanics and/or a decreased pulmonary capillary blood volume in 16 (73 percent) of the diabetic group. Since pulmonary dysfunction was either an isolated non-endocrine &nding or was associated with only early systemic complications in these young patients, our &ndings suggest that pulmonary dysfunction is an early measurable complication in insulin-dependent diabetes mellitus.

function in subjects with diabetes mellitus has, to date, been described in three previous clinical studies, the results of which are contradictory. One study involving 11 insulin-dependent diabetic subjects showed reduced elastic recoil at low lung volumes, associated with a reduced total lung capacity;1 a second study involving 20 insulin-dependent diabetic subjects showed no pulmonary abnormalities; 2 while a third study involving 7 4insulin-dependent diabetic subjects showed no abnormality of ventilatory mechanics, but did show a diminished transfer of carbon monoxide at the alveolar capillary level. 3 Recently it has been reported that young patients with insulin-dependent diabetes and limited joint mobility had smaller lung volumes than diabetic patients without limited joint mobility. 4 Since these authors did not measure lung or chest wall elastic recoil, they were unable to specify the cause of the restrictive defect. Histologic evidence of pulmonary abnormalities in streptozotocin-induced diabetes in rats have included alterations in the ultrastructure of granular pneumocytes in the inter-alveolar septum, 5 of non-ciliated bronchiolar epithelial (Clara) cells6 and of collagen and elastin in the alveolar wall. 7 It is possible that some of these ultrastructural abnormalities could be the result of direct toxic effects of streptozotocin, although Kida et al7 argued that they were the consequence of a lack

of insulin. Autopsy findings in diabetic subjects, however, have included thickening of epithelial and capillary basal laminae of alveoli, 8 and centrilobular emphysema. 9 Evidence has also been presented of diabetic micro-angiopathy in the capillaries of alveolar septa and also in the alveolar and pleural arterioles. 9 In view of the pathologic evidence and the contradictory clinical findings, the primary aim of this study was to assess lung mechanics and pulmonary gas transfer (both membrane diffusing capacity and pulmonary capillary blood volume) in young insulindependent diabetics. A secondary aim was to relate any dysfunction (if detected) to other diabetic complications.

~ng

•From the Endocrine Unit, Departments of Internal Medicine and Medical Physiology, University of Stellenbosch; and Tygerberg Hospital, Cape Town, Republic of South Africa. tSupported by the Medical Research Council ofSouth Africa. Part of this work is to be included in a Ph.D. thesis to be submitted to the University of Stellenbosch. iMRC Unit fur the Diffuse Obstructive Pulmonary Syndrome. Manuscript received March 3; revision accepted May 22. Reprint requests: Dr. Sandi.er, Department of Internal Medicine, Unit>erlUy of Stellenbosch, PO Box 63, 75fJ5 JYgerberg, South Africa.

870

MATERIALS AND METHODS

Subjects

Twenty-two insulin-dependent (type I) diabetic patients were investigated. All patients were dependent on insulin from the time of diagnosis, regularly attended the diabetic clinic and none had been admitted to hospital in the six months preceding the study. None of the subjects had ever smoked. Patients with a positive history of respiratory symptoms or disease, JO or who exhibited clinical or radiologic abnonnalitiesio of the cardiorespiratory system were excluded from the study. None ofthe patients had a history of allergic disorders and neither anti-nuclear nor rheumatoid fiictor was present in their serum. Each patient was clinically assessed fur the presence of cardiovascular, renal, retinal and neurologic alterations of diabetes. All patients were examined by an ophthalmologist to detect retinopathy, and other cardiovascular complications were assessed by chest radiographic and electrocardiographic examination. Glycosy1ated hemoglobin (HbAJ concentration (lsolab Corp.) was used as an index of diabetic control over the three-month period prior to the study. Renal dysfunction was gauged from the measurement of serum urea, creatinine and P2-microglobulin concentrations (Pharmacia Diagnostics) and also from the creatinine clearance and 24-hr urinary protein excretion. The presence of autonomic neuropathy was P\llmonary Funcllon In Young Olabalic Subjects (Sandlet;

Bunn, sr-t)

evaluated at the bedside in accordance with the guidelines of Ewing and Clarke11 and subclinical peripheral neuropathy was assessed by electroneurography. The insulin-dependent diabetic patients were compared with reference (non-diabetic) subjects, matched fur age, sex and race who were healthy, non-smoking normoglycemic volunteers free from respiratory symptoms or disease. The age of the patients, and nondiabetic subjects, ranged between 15 and 30 years with a mean ofl9. l and a standard deviation of5.0 years. The duration of diabetes in the patients varied from two months to 25 years, with a mean duration of 4. 75±5.50 years. All subjects voluntarily consented to the procedures after the nature of the tests had been explained to them. The project was approved by the faculty ethics committee.

Methods Lung mechanics were evaluated in patients and reference subjects from maximal expiratory flow-volume loops, the single-breath nitrogen washout test and static pulmonary elastic recoil. Flow-volume loops were recorded from a wedge spirometer (Med Science Corp). A minimum of three successive loops was obtained from each subject and the flow-volume curve with the greatest sum ofFEV1 and FVC was selected fur analysis. The single-breath nitrogen (SBN) washout tests were recorded immediately after the flow-volume loops. Washout test results were considered technically acceptable if inspired and expired slow vital capacities were within 5 percent of each other and agreed with the inspired capacity from the best flowvolume curve ( ± 5 percent). Furthermore, expired flow was kept between 300 and 500 ml/s by technologist and patient observing a graphic, real-time display of the flow rate plotted against volume on which scale the inspired vital capacity, obtained during the flowvolume test, was identified. Expired nitrogen concentration was measured with a calibrated rapid-response nitrogen analyzer (Med Science Corp) and was plotted against expired volume. The spirometer and nitrogen analyzer were linked to a microcomputer with graphics facilities (Tektronix 4052). Volume and flow data were expressed in BTPS units. The onset of phase 4 of the single breath nitrogen washout was determined by using a computer algorithm which employed linear regression analysis of phases 3 and 4: the point of intercept of these two lines indicates the closing volume. (The position of the closing point was checked "manually" and when there was a discrepancy of greater than 5 percent, the manually determined closing point was used.) Total lung capacity was determined from the single breath nitrogen washout test using the method described by Briindler and Lewis;11 expired volume of nitrogen was determined by digital integration of the area under the expired nitrogen volume curve. The flow-volume data, influenced by height and weight, were also expressed as a percentage of predicted normal values obtained from the regression equations of Schoenberg et al13 and Grimby and Soderholm. 14 Closing volume/vital capacity and closing capacity/ total lung capacity were also compared with the normal values of Buist and Ross15 (CVNC% and CCrrLC%). Lung elasticity was measured under static conditions in the diabetic group and in 11 control subjects by means of the esophageal balloon technique. 18 By immersing all but the very top of the balloon in water, the optimal balloon volume was determined to be 0.3 ml. The 10 cm long latex balloon (0.05-0.07 mm thick and 3.5 cm circumference) was then passed via the nose into the esophagus and positioned at the point of most negative pressure and minimum cardiac artifact. The lung volume was measured on the spirometer while transpulmonary pressure was simultaneously measured (Statham 13l'IC differential pressure transducer) under static conditions during interrupted passive deflation from TLC to Rv. A constant volume history was ensured by instructing patients to take five slow vital capacity breaths prior to duplicate pressure-volume measurements, from which all data points (at least eight per curve) were used fur analysis. The lung elastic recoil was expressed in terms

of the volume independent index, Kst(L), derived from a linear regression best fit of the data to the exponential equation described by Colebatch et al. 17 (V=A-Be-u, where V=volume, P=transpulmonary pressure and A, Band Kare constants). Values ofKst(L) were also compared with the reference values of the same authors. 17 The single-breath carbon monoxide transfer factor (TLCO) and its components were measured at low oxygen (F101 = 0.20) and high oxygen (F101 =0.80) concentrations at the same time of day in all subjects. A Morgan Thmsfer-test Model C linked to a Data Dec microprocessor was used and transfer values determined according to furmulas established by Cotes. Ml TI.co was measured twice at a low oxygen concentration, ensuring that the inspired volume was within 5 percent of the inspired vital capacity obtained in the flow-volume loop. Thereafter, the patients breathed pure oxygen fur five min prior to TI.co measurements at the high oxygen concentration. In successive measurements, the breath hold time was held constant (IOs) and inspired vital capacity agreed within 5 percent. TI.co values were corrected fur back tension. 18 The rate of reaction of carbon monoxide with hemoglobin (9) and the values fur the membrane diffusing capacity (DM) and pulmonary capillary blood volume (Qc) were derived from the furmulae given by Cotes. 18 The transfer factor and its components (DM and Qc) were expressed per unit alveolar gas volume (TuVA, DMNAand QcN A), the latter derived by helium dilution during the transfer tests, as has been suggested by Werner. 18 Prior to CO transfer measurements, hemoglobin and carboxyhemoglobin concentrations were measured spectrophotometrically (Instrument Laboratories IL 282), with the CO-oximeter calibrated against a standard specimen prior to the measurements. In determining differences between the diabetic and non-diabetic subjects, the Student's t-test fur unpaired data was used. In view of the relatively small sample size, the non-parametric Mann Whitney U test was also used and, in all instances, the two tests gave similar results with regard to statistical significance.

REsur.:rs None of the patients had evidence of retinopathy or autonomic neuropathy. Nine patients exhibited peripheral neuropathy (subclinical electroneurographic signs only in seven); one of these also had nephropathy. None of the subjects had limited joint mobility. 4 Therapeutic control of the diabetes mellitus in the three months prior to the study is reflected by the mean HbA1 concentration of 11.6±2.8 percent (range in patients 7to17.4 percent). In our laboratory, HbA1 concentration is 6.5 ± 8.5 percent in normal, nondiabetic subjects. A random serum glucose concentration was within the normal range (3.4 to 6.4 mmoVL) in all the reference subjects, and none of them had glycosuria. In none of the subjects was the carboxyhemoglobin concentration greater than 2 percent; this value is compatible with non-smoking urban dwellers. Lung mechanics data are summarized in Tuble 1. Both lung volume and flow variables were within the normal range13•14 and there was no difference between the diabetic and reference group values. The mean CC/rLC of the diabetic group, however, was significantly greater (p<0.05) than that of the reference group. In the reference group, only one value of CC/I'LC was outside the normal range (defined by predicted: 15 CC/I'LC + 1.65 SD) while in the diabetic CHEST I 90 I 5 I NOVEMBER, 1988

871

Table I-Lung Mechanica in Diabetic and Control Subjects

IO

Diabetic subjects Control subjects TLC (1) TLC(%)* RVfl'LC FVC (%)* FEV/FVC (%) FEV1/FVC (%)* FEF,., (%)* FEF'IS (%)* Kst(L) In Kst(L) P(st)L_ (cm H 20) AN2(%/L) CVNC(%) CCfl'LC (%)

5.10::!: 1.40 110.2::!: 11.3 25.0::!:5.13 101.6::!:12.7 88.37::!:6.29 101.8::!:6.3 111.1::!:26.0 101.4::!:33.2 0.148::!:0.045 -1.955::!:0.301 28.5::!:8.8 0.72::!:0.40 8.7::!:5.5 31.4::!:6.8

5.29::!:1.29 112.2::!:13.0 24.66::!:2.75 101.8::!: 12.5 91.11 :!::6.34 104.4::!:5.6 110.9::!:23.4 108.0::!:27.7 0.118:t0.030t -2.192::!:0.185 32.3::!:8.8 0.74::!:0.43 6.8::!:2.6 27.2::!:2.9

TLC (%)

80

70

eo p<0.05 p<0.05

50 15

5

p<0.01

*Value expressed as a percentage of the predicted nonnal value. 13-15 tn=ll. TLC= total lung capacity; RV= residual volume; FVC =forced expiratory vital capacity; FEV/FVC =forced expiratory volume in l sec expressed as a percentage of FVC; FEF50 =forced expiratory flow at 50% of FVC; FEF'IS=forced expiratory flow at 75% of FVC; Kst(L) =volume independent index of lung elasticity; Pst(L),... = transpulmonary pressure at TLC; AN 2 =slope of phase 3 ofthe single breath nitrogen washout test; CVNC =ratio of closing volume to FVC; CCfl'LC =ratio of closing capacity to TLC.

group, seven subjects (32 percent) had values of CC/I'LC% outside this range. These differences could not be explained by differences in the RV/I'LC ratio since the values for this variable were normal and similar in the two groups. In the diabetic group, the mean Kst(L) was greater than the mean value for the healthy subjects. Kst(L) was significantly greater than the reference value17 in six (27 percent) of the subjects, and in half of the patients with an elevated Kst(L) the CC/fLC% was also elevated. The mean values for ln Kst(L) were also significantly different in the two groups but there was no difference between the values for transpulmonary pressure at TLC (P(st)L....). The different shapes of the pressure-volume curves of the diabetic and reference subjects are depicted in Figure 1. In all, evidence of abnormal lung mechanics was present in ten diabetic subjects (45 percent). Pulmonary gas transfer data are summarized in Tuble 2. The mean TilVA (the actual value and the value expressed as a percentage of predicted18) was lower in the diabetic group. In five subjects (23 percent), the TLIVA was abnormally reduced due to an associated low QctVA. In three subjects, QctVA alone was abnormally low, although the TLIVA was within normal limits. Hemoglobin concentration was similar in the two groups. Pulmonary function, incorporating both lung mechanics and gas transfer data, was abnormal in 16 of the 22 diabetic patients (ie, 73 percent) and in only one of 172

100

25

PL CcmH 20I

35

FIGURE 1. Representative lung pressure-volume curves from the mean Kst(L) values of the reference group (K=0.118), the diabetic group (K = 0.148) and from the six diabetic patients with abnonnally elevated values ofKst(L) (K=0.205).

the non-diabetic subjects. In Tuble 3, the pattern of pulmonary dysfunction in these patients and its relationship to the presence of other diabetic complications, age, sex and duration of diabetes is depicted. Seven of the nine insulin-dependent diabetic patients with neuropathy had measurable pulmonary dysfunction while the remaining nine of the 16 patients with measurable pulmonary dysfunction had no other diabetic complications. In the diabetic group, there was no difference in age, duration of diabetes and level of controi (HbAJ between those patients with and without abnormal pulmonary function. DISCUSSION

The results of this study indicate the presence of alterations in either lung mechanics and/or gas transfer in a large proportion (73 percent) of young, nonsmoking, insulin-dependent diabetic patients. Closing Capacity

The increased values of CC/I'LC found in seven diabetic patients could not be explained by differences in RV/fLC. Becklake and Permutt19 concluded that Table 2-Pulmonary Gas transfer in Diabetic and Control Subjects

TuVA (ml/min/mm HgfL) TuVA (% pred) DMlVA (ml/min/mm HgfL) QcNA(ml/L) Hb (gfdl)

Diabetic subjects

Control subjects

5.25::!:0.68* 85.5:!:14.4t 10.29::!:1.92 12.1::!:2.l* 13.9::!:2.7

5.61::!:0.57 96.7::!:8.4 10.11::!:2.13 14.2::!:3.3 13.9::!: 1.1

*O.Ol
elevated CCfl'LC and ~NJL values were indicative of early abnormalities of the small airways, but the precise interpretation of these values remains somewhat contentious .., The slope of phase 3 of the single breath nitrogen washout curve and the FEFso and FEF75, on the other hand, were normal and similar in the two groups; this evidence mitigates against the presence of clearcut, intrinsic airways disease. Although there was no correlation between CC/fLC and Kst(L) or ln K, three of the seven patients with elevated CCfl'LC values also had abnormal lung pressurevolume relationships (Tuble 3). This suggests that reduced lung elasticity may partly explain the elevated CC/fLC found in some of the diabetic patients of this study. The precise significance of isolated increases in CC/fLC has yet to be clearly resolved .., Since recurrent respiratory infections are known to occur in diabetic subjects, this was considered as a possible explanation in our patients. However, careful historytaking failed to reveal any evidence for this possibility in this young, diabetic group. In none of the previous clinical studies was the single breath nitrogen washout test evaluated, and this may explain why mild abnormalities in lung mechanics were sometimes not detected. 11.3 There is therefore no evidence of small airways disease in diabetes, but perhaps a further study of a wider age spectrum and larger population of diabetic patients may help to clarify the interpretation of the isolated elevation of CCfl'LC found in some diabetic patients.

Elastic Recoil

Although the elastic recoil pressure of the lung at TLC was similar in the two groups, the shape of the static pressure-volume curve was significantly different as reflected by the higher mean Kst(L) (p
Table 3-Pattem of Pulmonary Dyafunction in Diabetic Subjecta Age

Sex

Duration

ccm.c

K

In K

TuVA

QcNA

20 16 18 16 19 16 29 24 15 16 15 15 30 23 15 27 15 17 14 14 23 24

M M

3 1.5 4 5 0.2 3 5 0.5 0.5 2 4 6 0.5 14 1 25 2 4 1.5 5 17 10

27.91 23.70 23.22 22.39 26.90 24.97 31.70 43.61* 34.14* 44.47* 41.19* 28.50 39.92* 32.33 31.17* 42.27* 28.65 28.22 26.80 27.21 28.37 32.72

0.186* 0.158 0.164 0.126 0.137 0.187* 0.100 0.114 0.246* 0.186* 0.100 0.154 0.188 0.232* 0.095 0.192* 0.117 0.144 0.111 0.110 0.100 0.098

-1.682 -1.845 -1.808 -2.071 -1.988 -1.627 -2.303 -2.171 -1.402 -1.682 -2.302 -1.871 -1.671 -1.461 -2.354 -1.650 -2.146 -1.938 -2.198 -2.207 -2.303 -2.323

5.01 5.19 4.74* 3.39* 4.50* 6.15 4.44* 5.88 5.31 5.70 5.25 4.71* 5.80 5.04 6.12 6.21 4.89 4.89 5.61 5.91 5.52 5.31

12.28 9.57* 10.86* 9.38* 8.43* 10.57 8.89* 12.11 11.89* 14.00 11.63 10.93* 12.50 9.67* 16.00 16.00 12.64 13.17 13.90 13.16 13.20 14.40

F M

F M

F F F M M

F M M

F M

F M

F M

F F

Neuropathy

+ + +

+t + + +t

+t +t

*Abnormal value. tin these subjects, neuropathy was evident on clinical examination; in the oth~rs: only e~ectroneurographic ~~ence was present. Age and duration of diabetes are expressed in years, CCm.c in percent, TliVA m ml/min/mm Hg and QcNA m ml/L. CHEST I 90 I 5 I NOVEMBER, 11188

m

accelerated aging. Haber et al22 showed that Kst(L) was an index of alveolar size and, since the latter increases with age, the increased mean Kst(L) in the diabetic patients of this study supports the contention of accelerated aging in respect to the lungs. There was no correlation between TLIVA or QctV A and Kst(L) or In Kst(L), and it is therefore unlikely that a reduction of capillary blood volume could account for the slight loss of elastic recoil detected in some of the diabetic patients. A further possible explanation for the loss of elastic recoil that has yet to be explored in human beings is that chronic hyperglycemia may induce some qualitative change in surfactant function. There was, however, no evidence of this in a study on animals. 7 It seems likely that glycosylation of connective tissue proteins accounts for the loss of elastic recoil and reduced TLC in some diabetic patients. Determining whether these effects are really a form of premature aging or not will require further investigation.

Carbon Monoxide Transfer Factor Micro-angiopathy of pulmonary vessels has been reported in pathologic8 •9 and clinical studies. 3 Oulhen et al, 3 however, included smokers and patients with previous overt respiratory disease in their diabetic group. The findings of a reduced TLIVA and/or QctV A in eight diabetic patients of our study support these earlier observations. In three of the patients, a reduced QcN A co-existed with a normal TLIVA. This apparent anomaly is explained by DMNA values which were at the upper limit of normality. Although alveolar and capillary membrane thickening has been reported in insulin-dependent diabetics, 8 normal values for DMNA in the patients of his study mitigate against this as being functionally detectable in young subjects. In our diabetic patients, we were unable to confirm the inverse relationship between TUVA and Kst(L) found by Knudson et alao in middle-aged smokers and nonsmokers, nor was there a correlation between QctV A and Kst(L). Knudson et alao concluded that the decline in TUVA with age might be related to increased alveolar size which would increase the intra-alveolar diffusion distance. Butler and Kleinemanllll postulated that the age-related decrease was the result of a diminished capillary density/alveolar diameter ratio. Our finding of reduced CO diffusion due to a low pulmonary capillary blood volume may therefore either be a form of premature aging or due to pulmonary micro-angiopathy; the latter is supported by histologic evidence in other studies. 8 •9 The possibility exists that the increased concentration of glycosylated hemoglobin in diabetes might be associated with an increase in the CO-hemoglobin reaction rate (8). This would result in a spuriously low transfer factor and pulmonary capillary blood volume. 874

In the absence of data to test this hypothesis directly, we argued that ifelevated theta values were the cause of spuriously reduced TIJVA, there should have been a correlation between HbA1 concentration and TIJVA in the diabetic patients. We were unable to demonstrate any such correlation between HbA1 and TIJVA or QcN A values (or these values expressed as a percentage of predicted). Yet another consideration is that the P50 may be decreased in diabetes. 27 This increase in affinity for oxygen would tend to reduce theta and result in a spuriously elevated pulmonary capillary blood volume, which is contrary to our findings. Failure to correlate HbA1 and CO transfer data in the patients of this study suggests that even though hemoglobin glycosylation may effect theta, this does not significantly in8uence the CO-transfer factor. Pulmonary Dysfunction and Diabetic Complications

Although lung volumes, elasticity and transfer factor may be reduced and closing capacity increased in young diabetic patients, the pattern of pulmonary dysfunction is not uniform. This may re8ect either the relatively small sample size or, alternatively, the almost random pattern oforgan involvement which is a feature of diabetes mellitus. In earlier clinical studies, l-3 pulmonary dysfunction was not associated with age, duration of diabetes, or the presence of diabetic complications. More recently, reduced lung volumes were reported to be associated with limited joint mobility. 4 In our study, however, abnormal pulmonary function was detected in seven of the nine subjects with peripheral neuropathy. In five of these seven patients there was only subclinical, electroneurographic evidence of neuropathy. Nine patients with pulmonary dysfunction had no evidence of any systemic diabetic complications and none of the total group had evidence of limited joint mobility or other organ involvement, except one subject who had pulmonary dysfunction, neuropathy and nephropathy. Furthermore, the mean duration of diabetes in patients with abnormal lung function tests was only 4. 7 years. In view of the above it is suggested that pulmonary dysfunction may be one of the earliest measurable nonmetabolic alterations in some insulindependent diabetics. A longitudinal study could help identify a temporal pattern oflung involvement and its possible relationship to other organ involvement in diabetes. Clinical experience suggests that the pulmonary defects in insulin-dependent diabetic patients are insufficient to cause significant respiratory embarrassment. However, it is possible that the presence of diabetes-induced abnormalities in lung mechanics may predispose to the recurrent respiratory infections often seen in diabetic patients. Centrilobular emphysema has recently been reported in a post mortem Pulmonary Function In Young Diabetic: Subjects (Sancler, Bunn, Sl8watt)

study. 8 However, our finding of a decreased elastic recoil together with normal gas mixing (~N 2 in Tu.hie 1) suggest that alveolar wall destruction is not present at this stage. Finally, if the reduced QcrVA reflects the presence of pulmonary microangiopathy, this may be a contributory factor in the pathogenesis of unexplained acute non-congestive pulmonary edemam and shock lung28 sometimes round in insulin-dependent diabetics. The clinical implications of pulmonary dysfunction in diabetes, however, remain uncertain. CONCLUSIONS

It is concluded that mild dysfunction of lung mechanics and CO diffusion was present in 16 of the 22 young, non-smoking, insulin-dependent diabetic subjects of this study. This may reflect premature aging of the lung. In contrast, lung function defects were detected in only one of the 22 matched healthy subjects. Abnormal pulmonary function either coexisted with early peripheral neuropathy or preceded the onset of other systemic diabetic complications. ACKNOWLEDGMENTS: The authors thank Dr. R. A. D. Scott for his assistance and also the fullowing clinical technologists of the Pulmonary Function Laboratory, 'fygerberg Hospital: Messrs. J. C. Ellis (Head), J. Mouton, E. Jones and P. de Koc)(. REFERENCES

1 Schuyler MR, Niewoehner DE, Inkley SR, Kohn R. Abnormal lung elasticity in juvenile onset diabetes mellitus. Am Rev Respir Dis 1976; 113:37-41 2 Schernthaner G, Haber P, Kummer F, Ludwig H. Lung elasticity in juvenile onset diabetes mellitus. Am Rev Respir Dis 1977; 116:544-46 3 Oulhen Ph, Barthelemy L, Bellet-Barthas M, Barragon T. Contribution A L'etude de la function respiratoire du diabetique insulino-dependent. Rev Fr Mal Respir 1982; 10:213-24 4 Schnapf BM, Banks RA, Silverstein JH, Rosenbloom AL, Chesrown SE, Loughlin GM. Pulmonary function in insulin dependent diabetes mellitus with limited joint mobility. Am Rev Respir Dis 1984; 130:930-32 5 Plopper CG, Morishige WK. Alterations in granular (type ii) pneumocyte ultrastructure by streptowtocin-induced diabetes in the rat. Lab Invest 1978; 38:143-48 6 Plopper CG, Morishige WK. Alterations in the ultra-structure of non-ciliated bronchiolar epithelial (Clara) cells by streptozotocin-induced diabetes in rats. Am Rev Respir Dis 1979; 120:1137-43 7 Kida K, Utsuyoma M, Tukizawa 'I: Thurlbeck WM. Changes in lung morphologic features and elasticity caused by streptozotocin-induced Diabetes Mellitus in growing rats. Am Rev Respir Dis 1983; 128:125-31 8 Vracko R, Thorning D, Huang nv. Basal lamina of Alveolar Epithelium and capillaries. Quantitative changes with aging and in diabetes mellitus. Am Rev Respir Dis 1979; 120:973-83 9 Kodolova IM, Lysenko IY, Saltykov BB. Changes in the lungs in

diabetes mellitus. Arkh Patol 1982; 44:35-40 10 Ferris BG. Epidemiology standardization project. Am Rev Respir Dis 1978; 118:11-23, 89-110 11 Ewing DJ, Clarke BF. Diagnosis and management of diabetic autonomic neuropathy. Br Med J 1982; 2:916-18 12 Briindler JP, Lewis CM. Estimation of lung volumes from nitrogen washout curves. Bull Eur Physiopath Respir 1981; 18:281-89 13 Schoenberg JB, Beck GS, Bouhuys A. Growth and decay of pulmonary function in healthy blacks and whites. Respir Physiol 1978; 33:367-73 14 Grimby G, Soderholm B. Spirometric studies in normal subjects III: static lung volumes and maximal voluntary ventilation in adults with a note on physical fitness. Acta Med Scand 1963; 173:199-206 15 Buist AS, Ross BB. Predicted values fur closing volumes using a modified single breath nitrogen test. Am Rev Respir Dis 1973; 107:744-52 16 Cotes JE. Lung function: assessment and application in medicine. 4th ed. London: Blackwell Scientific, 1979, pp 230-50 17 Colebatch HGH, Greaves IA, Ng CKY. Exponential analysis of elastic recoil and aging in healthy males and females. J Appl Physiol 1979; 47:683-91 18 Werner F, Beneken Kolmer HH. The CO single breath transfer factor of the lung. Generally acceptable normal values. PBugers Arch 1982; 393:269-74 19 Becklake MR, Permutt S. Evaluation of tests oflung function fur "screening" fur early detection of chronic obstructive lung disease. In: Macklem P, Permutt P, eds. The lung in the transition between health and disease. New York: Marcel Dekker, 1979, pp 345-74 20 Knudson RJ, Bloom JW, Knudson DE, Kaltenborn WT. Subclinical effects of smoking. Physiological comparison of healthy middle-aged smokers and non-smokers and interrelationships of lung function measurements. Chest 1984; 85:20-29 21 McCuaig KE, Vessal S, Coppin K, Wiggs BJR, Dahlby R, Pare PD. Variability in measurements of pressure-volume curves in normal subjects. Am Rev Respir Dis 1985; 131:656-58 22 Haber PS, ColebatchHJH, NgCKY, Greaves IA. Alveolar size as a determinant of pulmonary distensibility in mammalian lungs. J Appl Physiol 1983; 54:837-45 23 Kohn RR, Schnider SL. Glycosylation of human collagen. Diabetes 1982; 31:47-51 24 Vogt BW, Schleicher ED, Wieland OH. e-Amino-Lysine-bound glucose in human tissues obtained at autopsy. Diabetes 1982; 31:1123-27 25 Hamlin CR, Kohn RR, Luschin JH. Apparent accelerated aging ofhuman collagen in diabetes mellitus. Diabetes 1975; 24:902-04 26 Butler C III, Kleineman J. Capillary density: alveolar diameter, a morphometric approach to ventilation and perfusion. Am Rev Respir Dis 1970; 102:886-94 27 Horvath SM, Jackson RL. Significance of P"° under stress. In: Loeppky JA, Riedesel MC, eds. Oxygen transport to human tissues. Amsterdam: North Holland: Elsevier 1982, pp 305-15 28 Sprung CI, Rackow EC, Fein IA. Pulmonary edema-a complication of diabetic keto-acidosis. Chest 1980; 77:687-88 29 Carrol P, Metz R. Adult respiratory distress syndrome complicating severely uncontrolled diabetes mellitus: report of nine cases and review of the literature. Diabetes Care 1982; 5:574-80

CHEST I 90 I 5 I NOVEMBER, 1988

875