Insulin Secretion, Glycosylated Haemoglobin and Islet Cell Antibodies in Cystic Fibrosis Children and Adolescents with Different Degrees of Glucose Tolerance 1
1
2
7
2
Summary In comparison with 12 weight-matched controls, 39 children and adolescents with cystic fibrosis (CF) showed higher fasting glycaemic levels and both delayed and enhanced blood glucose responses to OGTT. Glycaemic response was normal in 30/39 patients (76.9 %), impaired in other 7 cases (18%) and diabetic in the remnant two (5.1 %). Fasting insulin levels and total insulin output during OGTT did not differ in patients and controls, but insulin peak in CF group was delayed and sustained. In the whole C F series mean HbA1 c was higher than in controls but no difference was found between patients with normal and those with impaired glucose tolerance. Islet cell antibodies were absent in the entire C F group. In conclusion, our results confirm the raised prevalence in CF of glucose tolerance abnormalities, which do not seem to depend on autoimmune factor involvement. Delayed insulin response to OGTT can be considered a very early expression of beta cell impairment in the course of CF. In our experience HbA1c assay did not constitute a sensitive and specific screening test for detection of the C patients with glucose intolerance. Key words Cystic Fibrosis — Glucose Intolerance — Insulin Secretion — Glycosylated Haemoglobin — Islet Cell Antibodies — Beta Cell Impairment
Introduction Diabetes mellitus is a well known complication of cystic fibrosis (CF) whose prevalence has been reported as increasing during the last years (Rodman, Doershuk and Roland 1986) due to the raised longevity of C F patients {Nielsen and Scholtz 1982; Szaff, Hoiby and Flensborg 1983). It generally appears after the first decade of life and is characterized by a high incidence of microvascular complications, which is similar to that of diabetic subjects without C F (Sullivan and Denning 1989). The onset of diabetes in CF is frequently preceded by a picture of impaired glucose tolerance (IGT), which has been reported in as many as 40 % of patients and seems to be
Horm.metab. Res. 23 (1991)495-498 © Georg Thieme Verlag Stuttgart • New York
accompanied by progressive clinical deterioration (Finkelstein, Wielinski, Elliot, Warwick, Barbosa, Wu and Klein 1988). Recently some authors (Stutchfleld, O'Halloran, Teale, Isherwood, Smith and Heaf 1987) suggested that periodic measurements of glycosylated haemoglobin concentrations would be of value for determining the presence of IGT in CF patients and for monitoring its course, allowing earlier detection and treatment of diabetes. The high incidence of insulinopenia in C F was attributed for some time to pancreatic fibrosis and consequent disruption of the Langerhans islets (Handwerger, Roth, Gorden, Di Sant'Agnese, Carpeter and Georges 1969), but recent publications suggested that autoimmune factors may be implicated in its aetiology (Stutchfield, O'Halloran, Smith, Woodrow, Bottazzo and Haef 1988; Wilkin, Stutchfield, Smith and Heaf 1987). Aim of the present study was to assess in 39 CF children and adolescents with fasting euglycaemia: a) carbohydrate tolerance and pancreatic b-cell function by a standard oral glucose tolerance test (OGTT); b) the ability of glycosylated haemoglobin assay to discriminate different degrees of glucose tolerance; c) the eventual involvement of autoimmunity, as evaluated by the finding of serum islet cell antibodies (ICA). Subjects and Methods This study covered 39 CF children and adolescents (mean age 13.6 ± 4.7 years; range 5.5-22.2) attending the CF center of our University Hospital and followed by two of us (S.C.N. and C.S.). These patients had been selected to enter into the study on the basis of their familial history negative for diabetes mellitus and of repeatedly normal — i. e. < 115 mg/dl — plasma glucose values found on random assessments during the last year. Patients' body mass index ranged from 13.3 to 24.2 (X 17.7 ± 2.5). Shwachman score (Shwachman and Kulcazycki 1958), employed for the evaluation of their clinical status, ranged in the whole group from 20 to 96 (X 75.4 ± 16.1). At the time of the study none of the patients was taking drugs that could influence glucose tolerance and none had been receiving long term treatment with beta-lactam antibiotics and/or steroids during the last three months. Severe liver and/or kidney dysfunction as well as acute infections were excluded in every case. Nutritional status was evaluated as percentage of ideal weight for height, according to Waterlow criteria (Waterlow 1974) successively adapted (Sutphen 1985).
Received: 12 June 1990
Accepted: 12 Febr. 1991 after revision
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
F.DeLuca .T.Arrigo , S. Conti Nibali , C. Sferlazzas , A. Gigante , E. Di Cesare and D. Cucinotta Institutes of Pediatrics and of2 Internal Medicine, University of Messina, Messina, Italy
Horm. metab. Res. 23 (1991)
F. De Luca, T. Arrigo, S. Conti Nibali, C. Sferlazzas, A. Gigante, E. Di Cesare and D. Cucinotta
Fig. 1 Mean (±SD) plasma glucose and insulin responses to OGTT in 39 CF patients ( - • - ) and in 12 controls ( - A - ) . *2p < 0.05; * * < 0.005; * * * < 0.001.
Table 1 Plasma glucose (mg/dl) and insulin (|iU/ml) responses to OGTT, HbAic and clinical data of the two CF males with diabetic glucose tolerance. Patients
OGTT:
-10'
0'
30'
60'
90'
120'
280'
Area
Glucose
81
85
191
257
332
300
306
459
Insulin
33
33
32
34
49
61
69
80
Glucose
93
93
125
200
313
273
123
372
Insulin
13
10
20
32
39
53
40
63
A
B
(%)
Age (yrs)
Shwachman score
6.9
20.3
60
6.1
11.2
77
HbA1c
After an overnight fast, each patient and 12 healthy subjects (mean age 10.5 ± 1.5 years, range 7.8—13.1), selected as controls on the basis of their body mass index (18.0 ± 2.3, range 14.2-23) underwent a standard OGTT (1.75 g/kg body weight, maximum 75 g). Blood samples were taken at - 10,0, 30, 60,90,120 and 180 min after glucose load, for measurement of plasma glucose and insulin levels. On the blood sample taken at time 0 glycosylated haemoglobin HbAic and serum ICA were also assessed. For at least 3 days before the study a weight-maintaining diet, containing 50% of carbohydrates, was consumed by patients.
Serum ICA were determined by indirect immunofluorescence method on unfixed cryostat sections of blood group 0 human pancreata (Gleichmann and Bottazzo 1987) in all CF patients. The prevalence of ICA detected in our laboratory is 62.7% in 59 IDDM patients within the first 6 months of the disease, 28% in 140 their first degree relatives and 0 % in 13 7 healthy controls, i.e. superimposable to the one generally reported (Drell and Notkins 1987).
Plasma glucose was determined by means of the glucose-oxidase method (Hitachi Autoanalyzer-Boehringer Mannheim, Milan, Italy). On the plasma frozen at — 4 °C insulin concentrations of patients and controls were estimated by radioimmunoassay in 4 different runs, using commercial kits (Diagnostic Products Corporation, Los Angeles, USA; intra-assay and inter-assay coefficients of variation 4.6 and 7.1 respectively at 50 uU/ml) all from the same lot. Glucose and insulin areas beneath the curves were calculated according to the formula: 0.25 (fasting value) + 0.5 (30 minute value) + 0.75 (60 min value) + 0.5 (120 min value) + 0.25 (180 min value) (Haffner, Storn, Mazuda, Pugh and Patterson 1986 modified).
Statistical evaluation of the data was performed by means of the Student's T unpaired test, Pearson's x2 test and correlation analysis; two-tailed p values < 0.05 were considered significant. Data are expressed as mean (X) ± SD.
HbA1c was assessed in all the patients and in a control group (23 healthy subjects aged between 7.2 and 14.8 years, X 11.1+2.1; mean body mass index 19.3 ±2.2, range 14.9-24.9) by high pressure liquid chromatography with a fully automated instrument (Diamat, Bio-Rad Laboratories, Richmond, USA) and expressed as a percentage of the total haemoglobin. Normal range in our laboratories is 4—6%. Intra-assay and inter-assay coefficients of variation were respectively 5.0 + 0.45 and 8.2 ± 1.1 %.
Informed consent was always requested and obtained from all the subjects included in this study or their parents.
Results Plasma glucose and insulin responses to OGTT in CF group as a whole and in controls are illustrated in Figure 1. In comparison with controls, CF patients showed on the average significantly higher fasting glycaemic levels and both delayed (glycaemic peak at 30 min in all the controls and in only 5 patients and between 60 and 90 min in 34/39 patients, X2 = 31.4, p < 0.0005) and enhanced blood glucose responses (areas beneath the curves 299 + 65 vs 217 ± 1 9 mg/dl, 2p < 0.001). According to W H O criteria glycaemic response was considered as normal in 30/39 patients (76.9%), impaired in other 7 cases (18%) and diabetic in the remnant
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
496
Fig. 2 Mean (±SD) plasma insulin responses to OGTT in CF patients with normal ( - • - ) and with impaired glucose tolerance ( A-).
Horm. metab. Res. 23 (1991) 497
Fig. 3 HbA 1c individual values of three different subsets of subjects plotted against the respective means (continuous lines) and our laboratory reference range (broken lines).
two (5.1%). Clinical and metabolic data of the two diabetic patients are separately scheduled in Table 1. Fasting insulin levels were similar in patients and controls; after OGTT insulin area under the curve in CF group was not significantly different from the controls' one (137.6 ±52.6 vs 128.6 + 47.7 u,U/ml). Insulin peak, however, in CF series was delayed (at 30 min in 8/12 controls and in only 1 patient, %2 = 25.9, p < 0.0005; between 90 and 120 min in 26/39 patients and in none of controls,X2= 16.3, p < 0.0005; at 60 min in the remnant subjects) and sustained, insulin levels resulting significantly higher than in controls during the last period of OGTT (Fig. 1). Such a pattern of insulin secretion was visible even when patients were distinguished on the basis of their glucose tolerance, since a preserved and delayed insulin secretion was always observed in CF subjects with either normal, impaired (Fig. 2) or diabetic (Table 1) glucose tolerance. No significant differences were recorded between CF-NGT and CF-IGT concerning both mean baseline insulin levels and insulin areas under the curves (139.6 + 56.2 vs 135 ±32.7 uU/ml, respectively). Moreover mean age (13.5 ±4.9 vs 13.1+3.6 yrs), body mass index (17.6 ±2.2 vs 19.1 ±3.7), and Shwachman score (75.2 ± 16.8 vs 85.0 ± 7.1) were similar in NGT and IGT patients and no correlation was found in the entire CF group between these parameters and insulin or glucose areas. Nutritional status resulted absolutely normal in all the IGT patients, ranging from 91 to 121 (X±SD 105±11). The individual values of HbA1c in CF-NGT, CF-IGT and controls are depicted in Figure 3. Supranormal (i. e. above the upper limit of the reference range for our laboratory) HbA1c values were found in the two patients with diabetic glucose tolerance (Table 1) and in 4 of those with NGT, but in none of the ones with IGT. HbA1c concentrations in the entire CF series were inversely related to Shwachman score (r= — 0.33, p < 0.05) and directly to age (r = 0.35, p < 0.05). ICA were absent in sera of all the patients. Discussion Our results confirm the raised prevalence of carbohydrate metabolism disorders in CF: impaired glucose tolerance in 18 % and diabetic tolerance in another 5 % of our
patients, selected on the basis of repeatedly normal fasting plasma glucose levels. As far as beta cell function is concerned, it resulted globally preserved in all the patients, even in those with diabetic glucose tolerance: basal insulin levels and insulin areas during OGTT did not differ either between CF group as a whole and controls or between NGT and IGT patients. However, the pattern of insulin secretion in CF was different from the one observed in normal subjects in that insulin peak was almost constantly delayed, as already reported by others {Andersen, Game, Heilmann, Petersen, Petersen and Koch 1988; Hartling, Game, Binder, Heilmann, Petersen, Petersen and Koch 1988). Such abnormality was evident not only in the patients with pathological glucose tolerance, but also in a large number of those with apparently normal glucose tolerance. Consequently delayed insulin response to OGTT can be considered a very early expression of beta cell impairment in the course of CF, which may precede and eventually lead to development of a patent glucose intolerance. Our data confirm previous reports (Bistritzer, Sack, Eshkol and Katznelson 1983; Caiger, Frost, Bruce and Wilson 1987; Finkelstein et al. 1988) of higher HbA1c average concentrations in CF patients as a whole than in controls. This finding well reflects the overall increase of both basal and after OGTT blood glucose levels observed in our CF population. Nevertheless in our study population HbA1c resulted superimposable in NGT and IGT groups. Moreover individual HbAl c values of our IGT patients overlapped with the ones of NGT subjects in the normal value range and an overlap was also recorded in the supranormal value range between NGT patients and those with diabetic tolerance. In our experience, therefore, HbAic assay does not constitute a very sensitive and specific screening test for the detection of glucose intolerance in CF patients. On the other hand, the irreplaceable role of OGTT in the assessment of glucose intolerance in non CF patients has been already well established and discrepancies between HbAic levels and OGTT have been pointed out in subjects with mild abnormalities of carbohy-
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
Glucose Tolerance in Cystic Fibrosis
Horm. metab. Res. 23 (1991)
F. De Luca, T. Arrigo, S. Conti Nibali, C. Sferlazzas, A. Gigante, E. Di Cesare and D. Cucinotta
Geffner, M. E., B. M. Lippe, N. K. Maclaren, W. J. Riley: Role of autoimmunity in insulinopenia and carbohydrate derangements associated with cystic fibrosis. J.Pediatr. 112:419-421 (1988) Gleichmann, H., G. F. Bottazzo: Progress towards standardization of cytoplasmic islet cell-antibody assay. Diabetes 36: 378-384 (1987) Haffner, S. M., M. P. Storn, M. P. Mazuda, J. A. Pugh, J. K. Patterson: Hyperinsulinemia in a population at high risk for non-insulin-dependent diabetes mellitus. N. Engl. J. med. 315:220-224 (1986) Handwerger, S., J. Roth, P. Gorden, P. Di Sant'Agnese, D. F. Carpeter, P. Georges: Glucose intolerance in cystic fibrosis. N. Engl. J. Med. 281:451-461(1969) Harding, S. G., S. Game, C. Binder, C. Heilmann, W. Petersen, K. E. Petersen, C. Koch: Proinsulin, insulin and C-peptide in cystic fibrosis after an oral glucose tolerance test. Diab. Res. 7: 165—169 (1988) Nielsen, O. H., P. O. Scholtz: Cystic fibrosis in Denmark in the period 1945 — 1981: evaluation of centralized treatment. Acta Paediatr. Finally, there was no evidence of specific autoScand. 301:107-119(1982) immune phenomena in any patient, since ICA test was always Rodman, H. M., C. F. Doershuk, J. M. Roland.The interaction of 2 disnegative, even when a diabetic tolerance was already present. eases: diabetes mellitus and cystic fibrosis. Medicine 65: 389—397 On the basis of these data and in accordance with others (1986) (Geffner, Lippe, Maclaren and Riley 1988), we cannot support Shwachman, H., L. L. Kulcazycki: Longterm study of one hundred the view that autoimmune mechanisms similar to those obfive patients with cysticfibrosis.Am. J. Dis. Chid. 96:6-15 (1958) Stuchfield, P. R, S. O'Halloran, C. S. Smith, J. C. Woodrow, G. F. Botserved in type 1 DM are sometimes involved in the developtazzo, D. Heaf: HLA type, islet cell antibodies and glucose intolerment of glucose intolerance in CF patients. ance in cysticfibrosis.Arch. Dis. Child. 63:1234-1239 (1988) Stutchfield, P. R., S. O'Halloran, J. D. Teale, D. Isherwood, C S. Acknowledgements Smith, D. Heaf: Glycosylated haemoglobin and glucose intolerance in cysticfibrosis.Arch. Dis. Child. 62:805-810 (1987) The authors are grateful to Mrs. Mimma Risitano Sullivan, M. M., C. R. Denning: Diabetic microangiopathy in patients and co-workers for their skilled technical assistance. with cystic fibrosis. Pediatrics 84:642 - 647 (1989) Sutphen, J. L.: Growth as a measure of nutritional status. J. Pediatr. Gastroenterol. Nutr. 4:169-181 (1985) References Szaffi M., N. Hoiby, E. W. Flensborg: Frequent antibiotic therapy improves survival of cystic fibrosis patients with chronic pseudoAndersen, O., S. Game, C. Heilmann, K. E. Petersen, W. Petersen, C. monas aeruginosa infection. Acta Paediatr. Scand. 72: 651—657 Koch: Glucose tolerance and insulin receptor binding to mono(1983) cytes and erythrocytes in patients with cystic fibrosis. Acta PaediVerillo, A., A. De Teresa, R. Golia, V. Nunziata: The relationship beatr.Scand. 77:67-71 (1988) tween glycosylated haemoglobin levels and various degrees of gluBistritzer, 71, J. Sack, A. Eshkol, D. Katznelson: Haemoglobin Al and cose intolerance. Diabetologia24:391-396 (1983) pancreatic beta cell function in cystic fibrosis. Isr. J. Med. Sci. 19: Waterlow, J. C: Some aspects of chiildhood malnutrition as a public 600-603(1983) Bolli, G., P. Compagnucci, M. G. Cartechini, F. Santeusanio, C. Cirotto,health problem. Br. Med. J. 4:88-90 (1974) L. Scionti, P. Brunetti: HbAl in subjects with abnormal glucose Wilkin, T. J., P. R. Stutchfield, C. S. Smith, D. P. Heaf: Autoimmunity, diabetes and cysticfibrosis.Lancet II: 157(1987) tolerance but normal fasting plasma glucose. Diabetes 29: 272— 277(1980) Caiger, P., G. J. Frost, L. E. Bruce, S. G. F. Wilson: Cystic fibrosis and diabetes mellitus. Arch. Dis. Child. 62:101 -102 (1987) Requests for reprints should be addressed to: Drell, D. W., A. L. Notkins: Multiple immunological abnormalities in patients with type 1 (insulin-dependent) diabetes mellitus. DiabeProf. Filippo De Luca tologia 30:132-143(1987) Istituto di Clinica Pediatrica Finkelstein, S. M., C. L. Wielinski, G. R. Elliot, W. J. Warwick, J. Barbosa, S. Wu, D. J. Klein: Diabetes mellitus associated with cystic Policlinico Universitario 1-98100 Messina (Italy) fibrosis. J. Pediatr. 112:373 - 377 (1988) drate tolerance (Bolli, Compagnucci, Cartechini, Santeusanio, Cirotto, Scionti and Brunetti 1980; Verrillo, De Teresa, Golia and Nunziata 1983). Consequently we agree that, in a screening program for early detection of hyperglycemia in CF, HbA1 c may be useful only if associated with post-prandial blood glucose determination and, eventually, OGTT (Sullivan and Denning 1989). In our series HbA1c levels correlated positively and negatively with the patients' age and Shwachman score, respectively. These findings suggest that in CF the mild abnormalities in carbohydrate tolerance indicated by elevated HbA1c concentrations increase with age and are associated with progressive clinical deterioration, as also concluded by others (Finkelstein et al. 1988).
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
498
1
2
7
2
Summary In comparison with 12 weight-matched controls, 39 children and adolescents with cystic fibrosis (CF) showed higher fasting glycaemic levels and both delayed and enhanced blood glucose responses to OGTT. Glycaemic response was normal in 30/39 patients (76.9 %), impaired in other 7 cases (18%) and diabetic in the remnant two (5.1 %). Fasting insulin levels and total insulin output during OGTT did not differ in patients and controls, but insulin peak in CF group was delayed and sustained. In the whole C F series mean HbA1 c was higher than in controls but no difference was found between patients with normal and those with impaired glucose tolerance. Islet cell antibodies were absent in the entire C F group. In conclusion, our results confirm the raised prevalence in CF of glucose tolerance abnormalities, which do not seem to depend on autoimmune factor involvement. Delayed insulin response to OGTT can be considered a very early expression of beta cell impairment in the course of CF. In our experience HbA1c assay did not constitute a sensitive and specific screening test for detection of the C patients with glucose intolerance. Key words Cystic Fibrosis — Glucose Intolerance — Insulin Secretion — Glycosylated Haemoglobin — Islet Cell Antibodies — Beta Cell Impairment
Introduction Diabetes mellitus is a well known complication of cystic fibrosis (CF) whose prevalence has been reported as increasing during the last years (Rodman, Doershuk and Roland 1986) due to the raised longevity of C F patients {Nielsen and Scholtz 1982; Szaff, Hoiby and Flensborg 1983). It generally appears after the first decade of life and is characterized by a high incidence of microvascular complications, which is similar to that of diabetic subjects without C F (Sullivan and Denning 1989). The onset of diabetes in CF is frequently preceded by a picture of impaired glucose tolerance (IGT), which has been reported in as many as 40 % of patients and seems to be
Horm.metab. Res. 23 (1991)495-498 © Georg Thieme Verlag Stuttgart • New York
accompanied by progressive clinical deterioration (Finkelstein, Wielinski, Elliot, Warwick, Barbosa, Wu and Klein 1988). Recently some authors (Stutchfleld, O'Halloran, Teale, Isherwood, Smith and Heaf 1987) suggested that periodic measurements of glycosylated haemoglobin concentrations would be of value for determining the presence of IGT in CF patients and for monitoring its course, allowing earlier detection and treatment of diabetes. The high incidence of insulinopenia in C F was attributed for some time to pancreatic fibrosis and consequent disruption of the Langerhans islets (Handwerger, Roth, Gorden, Di Sant'Agnese, Carpeter and Georges 1969), but recent publications suggested that autoimmune factors may be implicated in its aetiology (Stutchfield, O'Halloran, Smith, Woodrow, Bottazzo and Haef 1988; Wilkin, Stutchfield, Smith and Heaf 1987). Aim of the present study was to assess in 39 CF children and adolescents with fasting euglycaemia: a) carbohydrate tolerance and pancreatic b-cell function by a standard oral glucose tolerance test (OGTT); b) the ability of glycosylated haemoglobin assay to discriminate different degrees of glucose tolerance; c) the eventual involvement of autoimmunity, as evaluated by the finding of serum islet cell antibodies (ICA). Subjects and Methods This study covered 39 CF children and adolescents (mean age 13.6 ± 4.7 years; range 5.5-22.2) attending the CF center of our University Hospital and followed by two of us (S.C.N. and C.S.). These patients had been selected to enter into the study on the basis of their familial history negative for diabetes mellitus and of repeatedly normal — i. e. < 115 mg/dl — plasma glucose values found on random assessments during the last year. Patients' body mass index ranged from 13.3 to 24.2 (X 17.7 ± 2.5). Shwachman score (Shwachman and Kulcazycki 1958), employed for the evaluation of their clinical status, ranged in the whole group from 20 to 96 (X 75.4 ± 16.1). At the time of the study none of the patients was taking drugs that could influence glucose tolerance and none had been receiving long term treatment with beta-lactam antibiotics and/or steroids during the last three months. Severe liver and/or kidney dysfunction as well as acute infections were excluded in every case. Nutritional status was evaluated as percentage of ideal weight for height, according to Waterlow criteria (Waterlow 1974) successively adapted (Sutphen 1985).
Received: 12 June 1990
Accepted: 12 Febr. 1991 after revision
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
F.DeLuca .T.Arrigo , S. Conti Nibali , C. Sferlazzas , A. Gigante , E. Di Cesare and D. Cucinotta Institutes of Pediatrics and of2 Internal Medicine, University of Messina, Messina, Italy
Horm. metab. Res. 23 (1991)
F. De Luca, T. Arrigo, S. Conti Nibali, C. Sferlazzas, A. Gigante, E. Di Cesare and D. Cucinotta
Fig. 1 Mean (±SD) plasma glucose and insulin responses to OGTT in 39 CF patients ( - • - ) and in 12 controls ( - A - ) . *2p < 0.05; * * < 0.005; * * * < 0.001.
Table 1 Plasma glucose (mg/dl) and insulin (|iU/ml) responses to OGTT, HbAic and clinical data of the two CF males with diabetic glucose tolerance. Patients
OGTT:
-10'
0'
30'
60'
90'
120'
280'
Area
Glucose
81
85
191
257
332
300
306
459
Insulin
33
33
32
34
49
61
69
80
Glucose
93
93
125
200
313
273
123
372
Insulin
13
10
20
32
39
53
40
63
A
B
(%)
Age (yrs)
Shwachman score
6.9
20.3
60
6.1
11.2
77
HbA1c
After an overnight fast, each patient and 12 healthy subjects (mean age 10.5 ± 1.5 years, range 7.8—13.1), selected as controls on the basis of their body mass index (18.0 ± 2.3, range 14.2-23) underwent a standard OGTT (1.75 g/kg body weight, maximum 75 g). Blood samples were taken at - 10,0, 30, 60,90,120 and 180 min after glucose load, for measurement of plasma glucose and insulin levels. On the blood sample taken at time 0 glycosylated haemoglobin HbAic and serum ICA were also assessed. For at least 3 days before the study a weight-maintaining diet, containing 50% of carbohydrates, was consumed by patients.
Serum ICA were determined by indirect immunofluorescence method on unfixed cryostat sections of blood group 0 human pancreata (Gleichmann and Bottazzo 1987) in all CF patients. The prevalence of ICA detected in our laboratory is 62.7% in 59 IDDM patients within the first 6 months of the disease, 28% in 140 their first degree relatives and 0 % in 13 7 healthy controls, i.e. superimposable to the one generally reported (Drell and Notkins 1987).
Plasma glucose was determined by means of the glucose-oxidase method (Hitachi Autoanalyzer-Boehringer Mannheim, Milan, Italy). On the plasma frozen at — 4 °C insulin concentrations of patients and controls were estimated by radioimmunoassay in 4 different runs, using commercial kits (Diagnostic Products Corporation, Los Angeles, USA; intra-assay and inter-assay coefficients of variation 4.6 and 7.1 respectively at 50 uU/ml) all from the same lot. Glucose and insulin areas beneath the curves were calculated according to the formula: 0.25 (fasting value) + 0.5 (30 minute value) + 0.75 (60 min value) + 0.5 (120 min value) + 0.25 (180 min value) (Haffner, Storn, Mazuda, Pugh and Patterson 1986 modified).
Statistical evaluation of the data was performed by means of the Student's T unpaired test, Pearson's x2 test and correlation analysis; two-tailed p values < 0.05 were considered significant. Data are expressed as mean (X) ± SD.
HbA1c was assessed in all the patients and in a control group (23 healthy subjects aged between 7.2 and 14.8 years, X 11.1+2.1; mean body mass index 19.3 ±2.2, range 14.9-24.9) by high pressure liquid chromatography with a fully automated instrument (Diamat, Bio-Rad Laboratories, Richmond, USA) and expressed as a percentage of the total haemoglobin. Normal range in our laboratories is 4—6%. Intra-assay and inter-assay coefficients of variation were respectively 5.0 + 0.45 and 8.2 ± 1.1 %.
Informed consent was always requested and obtained from all the subjects included in this study or their parents.
Results Plasma glucose and insulin responses to OGTT in CF group as a whole and in controls are illustrated in Figure 1. In comparison with controls, CF patients showed on the average significantly higher fasting glycaemic levels and both delayed (glycaemic peak at 30 min in all the controls and in only 5 patients and between 60 and 90 min in 34/39 patients, X2 = 31.4, p < 0.0005) and enhanced blood glucose responses (areas beneath the curves 299 + 65 vs 217 ± 1 9 mg/dl, 2p < 0.001). According to W H O criteria glycaemic response was considered as normal in 30/39 patients (76.9%), impaired in other 7 cases (18%) and diabetic in the remnant
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
496
Fig. 2 Mean (±SD) plasma insulin responses to OGTT in CF patients with normal ( - • - ) and with impaired glucose tolerance ( A-).
Horm. metab. Res. 23 (1991) 497
Fig. 3 HbA 1c individual values of three different subsets of subjects plotted against the respective means (continuous lines) and our laboratory reference range (broken lines).
two (5.1%). Clinical and metabolic data of the two diabetic patients are separately scheduled in Table 1. Fasting insulin levels were similar in patients and controls; after OGTT insulin area under the curve in CF group was not significantly different from the controls' one (137.6 ±52.6 vs 128.6 + 47.7 u,U/ml). Insulin peak, however, in CF series was delayed (at 30 min in 8/12 controls and in only 1 patient, %2 = 25.9, p < 0.0005; between 90 and 120 min in 26/39 patients and in none of controls,X2= 16.3, p < 0.0005; at 60 min in the remnant subjects) and sustained, insulin levels resulting significantly higher than in controls during the last period of OGTT (Fig. 1). Such a pattern of insulin secretion was visible even when patients were distinguished on the basis of their glucose tolerance, since a preserved and delayed insulin secretion was always observed in CF subjects with either normal, impaired (Fig. 2) or diabetic (Table 1) glucose tolerance. No significant differences were recorded between CF-NGT and CF-IGT concerning both mean baseline insulin levels and insulin areas under the curves (139.6 + 56.2 vs 135 ±32.7 uU/ml, respectively). Moreover mean age (13.5 ±4.9 vs 13.1+3.6 yrs), body mass index (17.6 ±2.2 vs 19.1 ±3.7), and Shwachman score (75.2 ± 16.8 vs 85.0 ± 7.1) were similar in NGT and IGT patients and no correlation was found in the entire CF group between these parameters and insulin or glucose areas. Nutritional status resulted absolutely normal in all the IGT patients, ranging from 91 to 121 (X±SD 105±11). The individual values of HbA1c in CF-NGT, CF-IGT and controls are depicted in Figure 3. Supranormal (i. e. above the upper limit of the reference range for our laboratory) HbA1c values were found in the two patients with diabetic glucose tolerance (Table 1) and in 4 of those with NGT, but in none of the ones with IGT. HbA1c concentrations in the entire CF series were inversely related to Shwachman score (r= — 0.33, p < 0.05) and directly to age (r = 0.35, p < 0.05). ICA were absent in sera of all the patients. Discussion Our results confirm the raised prevalence of carbohydrate metabolism disorders in CF: impaired glucose tolerance in 18 % and diabetic tolerance in another 5 % of our
patients, selected on the basis of repeatedly normal fasting plasma glucose levels. As far as beta cell function is concerned, it resulted globally preserved in all the patients, even in those with diabetic glucose tolerance: basal insulin levels and insulin areas during OGTT did not differ either between CF group as a whole and controls or between NGT and IGT patients. However, the pattern of insulin secretion in CF was different from the one observed in normal subjects in that insulin peak was almost constantly delayed, as already reported by others {Andersen, Game, Heilmann, Petersen, Petersen and Koch 1988; Hartling, Game, Binder, Heilmann, Petersen, Petersen and Koch 1988). Such abnormality was evident not only in the patients with pathological glucose tolerance, but also in a large number of those with apparently normal glucose tolerance. Consequently delayed insulin response to OGTT can be considered a very early expression of beta cell impairment in the course of CF, which may precede and eventually lead to development of a patent glucose intolerance. Our data confirm previous reports (Bistritzer, Sack, Eshkol and Katznelson 1983; Caiger, Frost, Bruce and Wilson 1987; Finkelstein et al. 1988) of higher HbA1c average concentrations in CF patients as a whole than in controls. This finding well reflects the overall increase of both basal and after OGTT blood glucose levels observed in our CF population. Nevertheless in our study population HbA1c resulted superimposable in NGT and IGT groups. Moreover individual HbAl c values of our IGT patients overlapped with the ones of NGT subjects in the normal value range and an overlap was also recorded in the supranormal value range between NGT patients and those with diabetic tolerance. In our experience, therefore, HbAic assay does not constitute a very sensitive and specific screening test for the detection of glucose intolerance in CF patients. On the other hand, the irreplaceable role of OGTT in the assessment of glucose intolerance in non CF patients has been already well established and discrepancies between HbAic levels and OGTT have been pointed out in subjects with mild abnormalities of carbohy-
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
Glucose Tolerance in Cystic Fibrosis
Horm. metab. Res. 23 (1991)
F. De Luca, T. Arrigo, S. Conti Nibali, C. Sferlazzas, A. Gigante, E. Di Cesare and D. Cucinotta
Geffner, M. E., B. M. Lippe, N. K. Maclaren, W. J. Riley: Role of autoimmunity in insulinopenia and carbohydrate derangements associated with cystic fibrosis. J.Pediatr. 112:419-421 (1988) Gleichmann, H., G. F. Bottazzo: Progress towards standardization of cytoplasmic islet cell-antibody assay. Diabetes 36: 378-384 (1987) Haffner, S. M., M. P. Storn, M. P. Mazuda, J. A. Pugh, J. K. Patterson: Hyperinsulinemia in a population at high risk for non-insulin-dependent diabetes mellitus. N. Engl. J. med. 315:220-224 (1986) Handwerger, S., J. Roth, P. Gorden, P. Di Sant'Agnese, D. F. Carpeter, P. Georges: Glucose intolerance in cystic fibrosis. N. Engl. J. Med. 281:451-461(1969) Harding, S. G., S. Game, C. Binder, C. Heilmann, W. Petersen, K. E. Petersen, C. Koch: Proinsulin, insulin and C-peptide in cystic fibrosis after an oral glucose tolerance test. Diab. Res. 7: 165—169 (1988) Nielsen, O. H., P. O. Scholtz: Cystic fibrosis in Denmark in the period 1945 — 1981: evaluation of centralized treatment. Acta Paediatr. Finally, there was no evidence of specific autoScand. 301:107-119(1982) immune phenomena in any patient, since ICA test was always Rodman, H. M., C. F. Doershuk, J. M. Roland.The interaction of 2 disnegative, even when a diabetic tolerance was already present. eases: diabetes mellitus and cystic fibrosis. Medicine 65: 389—397 On the basis of these data and in accordance with others (1986) (Geffner, Lippe, Maclaren and Riley 1988), we cannot support Shwachman, H., L. L. Kulcazycki: Longterm study of one hundred the view that autoimmune mechanisms similar to those obfive patients with cysticfibrosis.Am. J. Dis. Chid. 96:6-15 (1958) Stuchfield, P. R, S. O'Halloran, C. S. Smith, J. C. Woodrow, G. F. Botserved in type 1 DM are sometimes involved in the developtazzo, D. Heaf: HLA type, islet cell antibodies and glucose intolerment of glucose intolerance in CF patients. ance in cysticfibrosis.Arch. Dis. Child. 63:1234-1239 (1988) Stutchfield, P. R., S. O'Halloran, J. D. Teale, D. Isherwood, C S. Acknowledgements Smith, D. Heaf: Glycosylated haemoglobin and glucose intolerance in cysticfibrosis.Arch. Dis. Child. 62:805-810 (1987) The authors are grateful to Mrs. Mimma Risitano Sullivan, M. M., C. R. Denning: Diabetic microangiopathy in patients and co-workers for their skilled technical assistance. with cystic fibrosis. Pediatrics 84:642 - 647 (1989) Sutphen, J. L.: Growth as a measure of nutritional status. J. Pediatr. Gastroenterol. Nutr. 4:169-181 (1985) References Szaffi M., N. Hoiby, E. W. Flensborg: Frequent antibiotic therapy improves survival of cystic fibrosis patients with chronic pseudoAndersen, O., S. Game, C. Heilmann, K. E. Petersen, W. Petersen, C. monas aeruginosa infection. Acta Paediatr. Scand. 72: 651—657 Koch: Glucose tolerance and insulin receptor binding to mono(1983) cytes and erythrocytes in patients with cystic fibrosis. Acta PaediVerillo, A., A. De Teresa, R. Golia, V. Nunziata: The relationship beatr.Scand. 77:67-71 (1988) tween glycosylated haemoglobin levels and various degrees of gluBistritzer, 71, J. Sack, A. Eshkol, D. Katznelson: Haemoglobin Al and cose intolerance. Diabetologia24:391-396 (1983) pancreatic beta cell function in cystic fibrosis. Isr. J. Med. Sci. 19: Waterlow, J. C: Some aspects of chiildhood malnutrition as a public 600-603(1983) Bolli, G., P. Compagnucci, M. G. Cartechini, F. Santeusanio, C. Cirotto,health problem. Br. Med. J. 4:88-90 (1974) L. Scionti, P. Brunetti: HbAl in subjects with abnormal glucose Wilkin, T. J., P. R. Stutchfield, C. S. Smith, D. P. Heaf: Autoimmunity, diabetes and cysticfibrosis.Lancet II: 157(1987) tolerance but normal fasting plasma glucose. Diabetes 29: 272— 277(1980) Caiger, P., G. J. Frost, L. E. Bruce, S. G. F. Wilson: Cystic fibrosis and diabetes mellitus. Arch. Dis. Child. 62:101 -102 (1987) Requests for reprints should be addressed to: Drell, D. W., A. L. Notkins: Multiple immunological abnormalities in patients with type 1 (insulin-dependent) diabetes mellitus. DiabeProf. Filippo De Luca tologia 30:132-143(1987) Istituto di Clinica Pediatrica Finkelstein, S. M., C. L. Wielinski, G. R. Elliot, W. J. Warwick, J. Barbosa, S. Wu, D. J. Klein: Diabetes mellitus associated with cystic Policlinico Universitario 1-98100 Messina (Italy) fibrosis. J. Pediatr. 112:373 - 377 (1988) drate tolerance (Bolli, Compagnucci, Cartechini, Santeusanio, Cirotto, Scionti and Brunetti 1980; Verrillo, De Teresa, Golia and Nunziata 1983). Consequently we agree that, in a screening program for early detection of hyperglycemia in CF, HbA1 c may be useful only if associated with post-prandial blood glucose determination and, eventually, OGTT (Sullivan and Denning 1989). In our series HbA1c levels correlated positively and negatively with the patients' age and Shwachman score, respectively. These findings suggest that in CF the mild abnormalities in carbohydrate tolerance indicated by elevated HbA1c concentrations increase with age and are associated with progressive clinical deterioration, as also concluded by others (Finkelstein et al. 1988).
Downloaded by: University of Pennsylvania Libraries. Copyrighted material.
498
