- Email: [email protected]
Type 1 diabetes in insulin-treated adult-onset diabetic subjects
Diabetes Research and Clinical Practice 42 (1998) 49 – 53
Type 1 diabetes in insulin-treated adult-onset diabetic subjects Jinny Willis a,*, Russell Scott a, Laurie Brown a, Paul Zimmet b, Ian MacKay c, Merrill Rowley c b
a Lipid and Diabetes Research Group, Hagley Building, Christchurch Hospital, Christchurch, New Zealand International Diabetes Institute, Caulfield General Medical Centre, 260 Kooyong Road, Caulfield South 3162, Australia c The Centre for Molecular Biology and Medicine, Monash Uni6ersity, Clayton, Australia
Received 31 March 1998; received in revised form 7 May 1998; accepted 1 August 1998
Abstract The frequency of autoimmune features was compared in adult-onset diabetic subjects either requiring insulin treatment within 12 months of diagnosis or progressing to insulin therapy after a latency of at least 5 years. Adult-onset insulin-treated diabetic subjects were sampled from the population-based Canterbury Diabetes Registry (n =1580). There were 237 (15%) registrants who met the study criteria of age B 75 years at 1 January 1993, age at diagnosis of diabetes ]45 years and duration of diabetes between 5 and 15 years; 101 subjects commenced insulin 5–15 years after diagnosis (group 1) and 80 subjects commenced insulin within 1 year of diagnosis (group 2). C-peptide levels, islet cell antibodies (ICA) and antibodies against glutamic acid decarboxylase (anti-GAD) were determined in all individuals from group 1 (n =27) and group 2 (n= 23) who agreed to be recruited to the study. The group 1 and group 2 samples did not differ significantly in their demographic characteristics, nor were they different from the two groups from which they were drawn (mean age, 64.2 years; age at diagnosis, 53.5 years; duration of diabetes, 10.7 years; body mass index, 28.6 kg/m2). Overall, 12 of the 50 (24%) study subjects tested positive for anti-GAD; 43% (10) of group 2 subjects were anti-GAD positive compared with only 7.4% (2) of group 1 subjects (PB 0.01). Postprandial C-peptide levels were significantly lower in group 2 subjects compared with group 1 subjects (627 vs 1124 pM, PB0.05). All subjects were ICA negative. These observations suggest that autoimmune destruction of b-cells explains early requirement for insulin in adult-onset diabetes. © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. Keywords: Adult-onset diabetes mellitus; Anti-GAD; Insulin-treated diabetes mellitus; Islet cell antibodies; Latent autoimmune diabetes in adults; Type 1 diabetes mellitus
* Corresponding author. Tel.: +64 3 3640448; fax: + 64 3 3640457; e-mail: [email protected] 0168-8227/98/$ - see front matter © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. PII: S0168-8227(98)00090-4
50
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53
1. Introduction Type 1 diabetes mellitus (formerly insulin-dependent diabetes mellitus, or IDDM) may present at any age. Incidence data for Type 1 diabetes in adults shows considerable variation in rates of presentation for different age groups. A cross-sectional study in East Finland revealed that, while most newly diagnosed Type 1 diabetes cases were in children and adolescents, a high proportion (37%) were diagnosed after the age of 19 years, with the disorder presenting very infrequently in subjects aged over 55 years [1]. Other studies have suggested incidence rates of Type 1 diabetes in adults aged 50–70 years virtually equal to those seen in subjects aged B20 years [2,3]. In our population (Canterbury, New Zealand) the incidence rates for adult-onset diabetic subjects who commenced insulin within 12 months of diagnosis were similar to those found for Type 1 diabetes in childhood and adolescence. Approximately 15% of all adult-onset diabetic subjects use insulin, and 85% of these insulin-treated cases have commenced insulin therapy within 1 year of diagnosis [2]. This implies an insulin deficiency syndrome which could be a consequence of autoimmune b-cell destruction, of either acute or delayed onset (LADA). Autoimmune markers of b-cell destruction, such as anti-glutamic acid decarboxylase (antiGAD) and islet cell antibodies (ICA), have proved useful predictors of diabetes type in adultonset diabetes mellitus [4,5]. Biochemical characteristics such as antibodies against b-cell antigens and b-cell products should be able to predict those who will proceed to insulin therapy, defining an important subgroup within adult-onset diabetes mellitus. The current study was undertaken to test the hypothesis that adults requiring insulin therapy within 12 months of diagnosis of diabetes mellitus have an autoimmune b-cell destructive process, in contrast to those converting to insulin after a longer interval post-diagnosis. This study determined the frequency of Type 1 diabetes, or latent autoimmune diabetes in adults (LADA), by measuring anti-GAD and ICA, and the levels of postprandial C-peptide, in a population-based
sample of insulin-treated adult-onset diabetic subjects drawn from the Canterbury Diabetes Register who commenced insulin therapy within 1 year of diagnosis or after a latency of 5 or more years [6].
2. Materials and methods Adult-onset insulin-treated diabetic subjects were sampled from the population-based Canterbury Diabetes Registry of insulin-treated diabetic individuals [6]. The study criteria of age B 75 years at 1 January 1993, age at diagnosis of diabetes ] 45 years and duration of diabetes of 5–15 years were met by 237 (15%) registrants. All of the 101 subjects who commenced insulin 5–15 years after diagnosis (group 1) and the 80 subjects who commenced insulin within 1 year of diagnosis (group 2) were advised of the study by mail and invited to participate. Postprandial blood samples were obtained from all individuals from group 1 (n= 27) and group 2 (n= 23) who elected to be recruited to the study, for determination of Cpeptide, ICA and anti-GAD. ICA were determined by indirect immunofluorescence on cryostat sections (5 mm) of human pancreas from a blood group O donor on aminoalkylsilane-coated slides [7]. Tissue was preincubated with ICA negative normal serum and serum samples were diluted with phosphatebuffered saline containing aprotonin (Bayer) to inhibit proteolytic activity during the overnight incubation at room temperature. The sections were incubated with fluorescein-labelled anti-human IgG (Atlantic Antibodies). Results were expressed in Juvenile Diabetes Foundation (JDF) units and a positive result was defined as ] 10 JDFU. The Tenth International Proficiency Programme showed our assay to have a sensitivity of 80% and a specificity of 100%. Anti-GAD was determined by radioimmunoprecipitation using 125 I-labelled porcine GAD with precipitation of 125 I-labelled GAD–antibody complexes by protein A Sepharose [8]. A positive result is defined as ] 18 U [9]. C-peptide was determined by radioimmunoassay as described previously [10].
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53
51
Table 1 Clinical and biochemical features of adult diabetic subjects
Age (years) Age at diagnosis (years) Duration of diabetes (years) Anti-GAD positive ICA positive C-peptide (pM) Body mass index (kg/m2)
Group 1 (n=27)
Group 2 (n = 23)
Significance (Student’s t-test)
64.8 (1.0) 53.2 (1.0) 11.6 (0.5) 2 0 1124 (172) 29.4 (1.3)
63.6 (1.4) 53.92 (1.2) 9.7 (0.7) 10 0 627 (133) 27.5 (1.1)
NS NS PB0.05 PB0.01 NS PB0.05 NS
Results are given as mean (S.E.M.). NS, not significant. Group 1 subjects commenced insulin therapy 5–15 years after diagnosis. Group 2 subjects commenced insulin therapy within 1 year of diagnosis.
3. Results The 50 patients had a mean (S.E.M.) age of 64.2 (0.9) years, age at diagnosis of 53.5 (0.8) years, and body mass index of 28.6 (0.9) kg/m2. Demographic and biochemical characteristics for the two groups are shown in Table 1. The group 1 individuals (commencing insulin 5 – 15 years after diagnosis) and group 2 individuals (commencing insulin within 1 year of diagnosis) did not differ significantly in age, or age at diagnosis, nor were they different from the two groups from which they were drawn. However, duration of diabetes was slightly longer in group 1 individuals, simply reflecting the age structure of those who were recruited to the study. Overall, 12 of the 50 (24%) study subjects tested positive for antiGAD; 43% (10) of group 2 subjects were antiGAD positive compared to only 7.4% (2) of group 1 subjects (PB0.01). Postprandial C-peptide levels were significantly lower in group 2 subjects compared with group 1 subjects. In addition, the mean C-peptide level in the 12 anti-GAD positive individuals was 116 pM (95% CI, 21–211 pM), compared with 1116 pM (95% CI, 859–1373 pM) for anti-GAD negative subjects (PB0.001). Fig. 1 illustrates the C-peptide concentrations in group 1 and group 2 individuals.
4. Discussion Clinical features have proved unreliable as a
basis for classifying autoimmune based insulin-dependency in newly presenting adult diabetic subjects [4]. For instance, Type 1 diabetes patients need not present with ketosis [11]. Furthermore, absolute insulin deficiency may not be present at the time of diagnosis as is typical for childhood and adolescent subjects. A prospective study of 41 consecutively referred newly diagnosed diabetic patients, aged 20–80 years, demonstrated that only 75% of patients could be correctly classified on the basis of age and percentage desirable body weight [12]. Insulin treatment is not a reliable indicator of absolute insulin deficiency. Patients
Fig. 1. Postprandial C-peptide levels measured 5 – 15 years after diagnosis in insulin-treated adult-onset diabetic patients requiring insulin therapy either within 12 months of diagnosis ( ) or 5 – 15 years after diagnosis ( ).
52
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53
with Type 2 diabetes may be given insulin as a treatment for symptomatic hyperglycaemia or for inadequate glycaemic control [13]. All subjects in this study were ICA negative. There are reports in the literature that ICA can be detected up to 10 years after diagnosis [11]. In childhood and adolescent Type 1 diabetes cases, ICA were persistently positive in 35% of subjects after 5 years duration of diabetes, with decline of ICA positivity being most rapid in abrupt-onset compared with slow-onset cases [14]. It is likely that, had the immune markers been measured early in the natural history of the disease in this cohort, the frequency of detection of anti-GAD, and particularly ICA, would have been increased. Earlier in the natural history, however, some patients may still have had appreciable b-cell function, whereas after a duration of 5–15 years, as expected, those with autoimmune expression showed low C-peptide levels implying minimal residual b-cell function. Whilst anti-GAD levels have been demonstrated to be constant with time [5], the small difference in duration of diabetes between the two patient groups potentially could have resulted in some group 1 individuals having seroconverted negatively prior to sampling. A recently published Finnish study in women aged 20 – 39 years revealed that anti-GAD was detected as early as 10 years prior to diagnosis of Type 1 diabetes and that there was little variation in the levels of anti-GAD during the prediabetes period [5]. In diabetic children, levels of anti-GAD at 3 years after diagnosis were similar to levels at diagnosis, but decreased by 6 – 7 years after diagnosis [15]. Thus, the possibility that the prevalence of anti-GAD in group 1 subjects may have been greater if measured at diagnosis, rather than after 5 – 15 years from diagnosis, cannot be discounted. Immune markers do not give absolute specificity for detection of Type 1 diabetes. The combination of immune markers, clinical decision to use insulin and depressed C-peptide levels, however, supports autoimmune b-cell destruction. Immune markers also provide the opportunity for detecting an ongoing process of b-cell de-
struction at a point in the natural history where C-peptide is not yet depressed and there are still insulin-producing cells left to preserve. These markers also allow for identifying at the clinical onset of the disease those individuals who will inevitably require insulin therapy. This could assist the clinician in marginal decisions, where the use of oral agents is likely to be poorly effective or, at best, of short-term benefit only. From these data obtained 5–15 years after diagnosis of diabetes mellitus, 43% of patients using insulin within 1 year of diagnosis would appear to have autoimmune-based b-cell destruction (Type 1 diabetes or latent expression of Type 1 diabetes, LADA), and could have been identified at presentation by measurement of antibodies directed against b-cells. In contrast, patients converting to insulin after a latency of more than 5 years very likely have b-cell loss of non-immune nature. The true prevalence of autoimmune-based insulin deficiency amongst this patient group is not known. A previous study by our group in newly diagnosed adult-onset diabetic patients found a prevalence of anti-GAD and ICA of almost 60% in cases categorised as insulin deficient on the basis of the insulin response to a mixed meal test [16]. Type 1 diabetes represents an important subset of adult-onset diabetes mellitus. Accurate classification at diagnosis is important as it could alter management, particularly as ethical interventions for prevention or delay of b-cell destruction are developed.
References [1] M. Laasko, K. Pyrola, Age of onset and type of diabetes, Diabetes Care 8 (1985) 114 – 117. [2] R.S. Scott, L.J. Brown, Prevalence and incidence of insulin-treated diabetes mellitus in adults in Canterbury, New Zealand, Diabetic Med. 8 (1991) 1 – 8. [3] A.G. Molbak, B. Christau, B. Marner, K. Borch-Johnsen, J. Nerup, Incidence of insulin-dependent diabetes mellitus in age groups over 30 years in Denmark, Diabetic Med. 11 (1994) 650 – 655. [4] L.C. Groop, G.F. Botazzo, D. Doniach, Islet cell antibodies identify latent type 1 diabetes in patients aged 35 – 75 years at diagnosis, Diabetes 35 (1986) 237 – 241.
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53 [5] J. Tuomilehto, P.Z. Zimmet, I.R. Mackay, et al., Antibodies to glutamic acid decarboxylase as predictors of insulin-dependent diabetes mellitus before clinical onset of disease, Lancet 343 (1994) 1383–1385. [6] L.J. Brown, R.S. Scott, A population based diabetes register — development and applications, Community Health Stud. 12 (1988) 437–441. [7] L.V. Forbes, L.J. Brown, R.S. Scott, B.A. Darlow, Immunogenetic, clinical and demographic characterisation of childhood Type 1 (insulin-dependent) diabetes in New Zealand, Diabetes Care 18 (1995) 1–8. [8] M.J. Rowley, I.R. Mackay, Q.-Y. Chen, W.J. Knowles, P.Z. Zimmet, Antibodies to glutamic acid decarboxylase discriminate major types of diabetes mellitus, Diabetes 41 (1992) 548 – 551. [9] Q.-Y. Chen, M.J. Rowley, G.C. Byrne, et al., Antibodies to glutamic acid decarboxylase in Australian children with insulin dependent diabetes mellitus and their first degree relatives, Pediatr. Res. 34 (1993) 785–790. [10] R.S. Scott, E.A. Espiner, R.A. Donald, M.J. Ellis, Free insulin, C-peptide and glucagon profiles in Type 1 diabetes, Aust. New Zealand J. Med. 10 (1980) 145–150. [11] A. Gottsater, U. Samuelsson, S. Nilsson, A. Lernmark, G. Sundkvist, Islet cell antibodies and fasting plasma
.
[12]
[13] [14]
[15]
[16]
53
C-peptide during the first 10 yr after diagnosis in patients with diabetes mellitus diagnosed in adult age, Diabetes Nutr. Metab. 5 (1992) 243 – 248. O. Hother-Nielsen, O. Faber, N. Schwatz Sorensen, H. Beck-Nielsen, Classification of newly diagnosed diabetic patients as insulin-requiring or non-insulin-requiring based on clinical and biochemical variables, Diabetes Care 11 (1988) 531 – 537. S. Madsbad, Classification of diabetes in older adults, Diabetes Care 13 (Suppl. 2) (1990) 93 – 96. T. Urakami, M. Owada, Y. Miyamoto, T. Kitagawa, H. Matsunaga, Serial changes in the prevalence of islet cell antibodies and islet cell antibody titer in children with Type 1 diabetes of abrupt or slow onset, Diabetes Care 18 (1995) 1095 – 1099. W.A. Hagopian, A.E. Karlsen, A. Gottsater, et al., Quantitative assay using recombinant human islet glutamic acid decarboxylase (GAD65) shows that 64K autoantibody positivity at onset predicts diabetes type, J. Clin. Invest. 91 (1993) 368 – 374. J.A. Willis, R.S. Scott, L.J. Brown, et al., Islet cell antibodies and antibodies against glutamic acid decarboxylase in newly diagnosed adult-onset diabetes mellitus, Diabetes Res. Clin. Pract. 33 (1996) 89 – 97.
Type 1 diabetes in insulin-treated adult-onset diabetic subjects Jinny Willis a,*, Russell Scott a, Laurie Brown a, Paul Zimmet b, Ian MacKay c, Merrill Rowley c b
a Lipid and Diabetes Research Group, Hagley Building, Christchurch Hospital, Christchurch, New Zealand International Diabetes Institute, Caulfield General Medical Centre, 260 Kooyong Road, Caulfield South 3162, Australia c The Centre for Molecular Biology and Medicine, Monash Uni6ersity, Clayton, Australia
Received 31 March 1998; received in revised form 7 May 1998; accepted 1 August 1998
Abstract The frequency of autoimmune features was compared in adult-onset diabetic subjects either requiring insulin treatment within 12 months of diagnosis or progressing to insulin therapy after a latency of at least 5 years. Adult-onset insulin-treated diabetic subjects were sampled from the population-based Canterbury Diabetes Registry (n =1580). There were 237 (15%) registrants who met the study criteria of age B 75 years at 1 January 1993, age at diagnosis of diabetes ]45 years and duration of diabetes between 5 and 15 years; 101 subjects commenced insulin 5–15 years after diagnosis (group 1) and 80 subjects commenced insulin within 1 year of diagnosis (group 2). C-peptide levels, islet cell antibodies (ICA) and antibodies against glutamic acid decarboxylase (anti-GAD) were determined in all individuals from group 1 (n =27) and group 2 (n= 23) who agreed to be recruited to the study. The group 1 and group 2 samples did not differ significantly in their demographic characteristics, nor were they different from the two groups from which they were drawn (mean age, 64.2 years; age at diagnosis, 53.5 years; duration of diabetes, 10.7 years; body mass index, 28.6 kg/m2). Overall, 12 of the 50 (24%) study subjects tested positive for anti-GAD; 43% (10) of group 2 subjects were anti-GAD positive compared with only 7.4% (2) of group 1 subjects (PB 0.01). Postprandial C-peptide levels were significantly lower in group 2 subjects compared with group 1 subjects (627 vs 1124 pM, PB0.05). All subjects were ICA negative. These observations suggest that autoimmune destruction of b-cells explains early requirement for insulin in adult-onset diabetes. © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. Keywords: Adult-onset diabetes mellitus; Anti-GAD; Insulin-treated diabetes mellitus; Islet cell antibodies; Latent autoimmune diabetes in adults; Type 1 diabetes mellitus
* Corresponding author. Tel.: +64 3 3640448; fax: + 64 3 3640457; e-mail: [email protected] 0168-8227/98/$ - see front matter © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. PII: S0168-8227(98)00090-4
50
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53
1. Introduction Type 1 diabetes mellitus (formerly insulin-dependent diabetes mellitus, or IDDM) may present at any age. Incidence data for Type 1 diabetes in adults shows considerable variation in rates of presentation for different age groups. A cross-sectional study in East Finland revealed that, while most newly diagnosed Type 1 diabetes cases were in children and adolescents, a high proportion (37%) were diagnosed after the age of 19 years, with the disorder presenting very infrequently in subjects aged over 55 years [1]. Other studies have suggested incidence rates of Type 1 diabetes in adults aged 50–70 years virtually equal to those seen in subjects aged B20 years [2,3]. In our population (Canterbury, New Zealand) the incidence rates for adult-onset diabetic subjects who commenced insulin within 12 months of diagnosis were similar to those found for Type 1 diabetes in childhood and adolescence. Approximately 15% of all adult-onset diabetic subjects use insulin, and 85% of these insulin-treated cases have commenced insulin therapy within 1 year of diagnosis [2]. This implies an insulin deficiency syndrome which could be a consequence of autoimmune b-cell destruction, of either acute or delayed onset (LADA). Autoimmune markers of b-cell destruction, such as anti-glutamic acid decarboxylase (antiGAD) and islet cell antibodies (ICA), have proved useful predictors of diabetes type in adultonset diabetes mellitus [4,5]. Biochemical characteristics such as antibodies against b-cell antigens and b-cell products should be able to predict those who will proceed to insulin therapy, defining an important subgroup within adult-onset diabetes mellitus. The current study was undertaken to test the hypothesis that adults requiring insulin therapy within 12 months of diagnosis of diabetes mellitus have an autoimmune b-cell destructive process, in contrast to those converting to insulin after a longer interval post-diagnosis. This study determined the frequency of Type 1 diabetes, or latent autoimmune diabetes in adults (LADA), by measuring anti-GAD and ICA, and the levels of postprandial C-peptide, in a population-based
sample of insulin-treated adult-onset diabetic subjects drawn from the Canterbury Diabetes Register who commenced insulin therapy within 1 year of diagnosis or after a latency of 5 or more years [6].
2. Materials and methods Adult-onset insulin-treated diabetic subjects were sampled from the population-based Canterbury Diabetes Registry of insulin-treated diabetic individuals [6]. The study criteria of age B 75 years at 1 January 1993, age at diagnosis of diabetes ] 45 years and duration of diabetes of 5–15 years were met by 237 (15%) registrants. All of the 101 subjects who commenced insulin 5–15 years after diagnosis (group 1) and the 80 subjects who commenced insulin within 1 year of diagnosis (group 2) were advised of the study by mail and invited to participate. Postprandial blood samples were obtained from all individuals from group 1 (n= 27) and group 2 (n= 23) who elected to be recruited to the study, for determination of Cpeptide, ICA and anti-GAD. ICA were determined by indirect immunofluorescence on cryostat sections (5 mm) of human pancreas from a blood group O donor on aminoalkylsilane-coated slides [7]. Tissue was preincubated with ICA negative normal serum and serum samples were diluted with phosphatebuffered saline containing aprotonin (Bayer) to inhibit proteolytic activity during the overnight incubation at room temperature. The sections were incubated with fluorescein-labelled anti-human IgG (Atlantic Antibodies). Results were expressed in Juvenile Diabetes Foundation (JDF) units and a positive result was defined as ] 10 JDFU. The Tenth International Proficiency Programme showed our assay to have a sensitivity of 80% and a specificity of 100%. Anti-GAD was determined by radioimmunoprecipitation using 125 I-labelled porcine GAD with precipitation of 125 I-labelled GAD–antibody complexes by protein A Sepharose [8]. A positive result is defined as ] 18 U [9]. C-peptide was determined by radioimmunoassay as described previously [10].
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53
51
Table 1 Clinical and biochemical features of adult diabetic subjects
Age (years) Age at diagnosis (years) Duration of diabetes (years) Anti-GAD positive ICA positive C-peptide (pM) Body mass index (kg/m2)
Group 1 (n=27)
Group 2 (n = 23)
Significance (Student’s t-test)
64.8 (1.0) 53.2 (1.0) 11.6 (0.5) 2 0 1124 (172) 29.4 (1.3)
63.6 (1.4) 53.92 (1.2) 9.7 (0.7) 10 0 627 (133) 27.5 (1.1)
NS NS PB0.05 PB0.01 NS PB0.05 NS
Results are given as mean (S.E.M.). NS, not significant. Group 1 subjects commenced insulin therapy 5–15 years after diagnosis. Group 2 subjects commenced insulin therapy within 1 year of diagnosis.
3. Results The 50 patients had a mean (S.E.M.) age of 64.2 (0.9) years, age at diagnosis of 53.5 (0.8) years, and body mass index of 28.6 (0.9) kg/m2. Demographic and biochemical characteristics for the two groups are shown in Table 1. The group 1 individuals (commencing insulin 5 – 15 years after diagnosis) and group 2 individuals (commencing insulin within 1 year of diagnosis) did not differ significantly in age, or age at diagnosis, nor were they different from the two groups from which they were drawn. However, duration of diabetes was slightly longer in group 1 individuals, simply reflecting the age structure of those who were recruited to the study. Overall, 12 of the 50 (24%) study subjects tested positive for antiGAD; 43% (10) of group 2 subjects were antiGAD positive compared to only 7.4% (2) of group 1 subjects (PB0.01). Postprandial C-peptide levels were significantly lower in group 2 subjects compared with group 1 subjects. In addition, the mean C-peptide level in the 12 anti-GAD positive individuals was 116 pM (95% CI, 21–211 pM), compared with 1116 pM (95% CI, 859–1373 pM) for anti-GAD negative subjects (PB0.001). Fig. 1 illustrates the C-peptide concentrations in group 1 and group 2 individuals.
4. Discussion Clinical features have proved unreliable as a
basis for classifying autoimmune based insulin-dependency in newly presenting adult diabetic subjects [4]. For instance, Type 1 diabetes patients need not present with ketosis [11]. Furthermore, absolute insulin deficiency may not be present at the time of diagnosis as is typical for childhood and adolescent subjects. A prospective study of 41 consecutively referred newly diagnosed diabetic patients, aged 20–80 years, demonstrated that only 75% of patients could be correctly classified on the basis of age and percentage desirable body weight [12]. Insulin treatment is not a reliable indicator of absolute insulin deficiency. Patients
Fig. 1. Postprandial C-peptide levels measured 5 – 15 years after diagnosis in insulin-treated adult-onset diabetic patients requiring insulin therapy either within 12 months of diagnosis ( ) or 5 – 15 years after diagnosis ( ).
52
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53
with Type 2 diabetes may be given insulin as a treatment for symptomatic hyperglycaemia or for inadequate glycaemic control [13]. All subjects in this study were ICA negative. There are reports in the literature that ICA can be detected up to 10 years after diagnosis [11]. In childhood and adolescent Type 1 diabetes cases, ICA were persistently positive in 35% of subjects after 5 years duration of diabetes, with decline of ICA positivity being most rapid in abrupt-onset compared with slow-onset cases [14]. It is likely that, had the immune markers been measured early in the natural history of the disease in this cohort, the frequency of detection of anti-GAD, and particularly ICA, would have been increased. Earlier in the natural history, however, some patients may still have had appreciable b-cell function, whereas after a duration of 5–15 years, as expected, those with autoimmune expression showed low C-peptide levels implying minimal residual b-cell function. Whilst anti-GAD levels have been demonstrated to be constant with time [5], the small difference in duration of diabetes between the two patient groups potentially could have resulted in some group 1 individuals having seroconverted negatively prior to sampling. A recently published Finnish study in women aged 20 – 39 years revealed that anti-GAD was detected as early as 10 years prior to diagnosis of Type 1 diabetes and that there was little variation in the levels of anti-GAD during the prediabetes period [5]. In diabetic children, levels of anti-GAD at 3 years after diagnosis were similar to levels at diagnosis, but decreased by 6 – 7 years after diagnosis [15]. Thus, the possibility that the prevalence of anti-GAD in group 1 subjects may have been greater if measured at diagnosis, rather than after 5 – 15 years from diagnosis, cannot be discounted. Immune markers do not give absolute specificity for detection of Type 1 diabetes. The combination of immune markers, clinical decision to use insulin and depressed C-peptide levels, however, supports autoimmune b-cell destruction. Immune markers also provide the opportunity for detecting an ongoing process of b-cell de-
struction at a point in the natural history where C-peptide is not yet depressed and there are still insulin-producing cells left to preserve. These markers also allow for identifying at the clinical onset of the disease those individuals who will inevitably require insulin therapy. This could assist the clinician in marginal decisions, where the use of oral agents is likely to be poorly effective or, at best, of short-term benefit only. From these data obtained 5–15 years after diagnosis of diabetes mellitus, 43% of patients using insulin within 1 year of diagnosis would appear to have autoimmune-based b-cell destruction (Type 1 diabetes or latent expression of Type 1 diabetes, LADA), and could have been identified at presentation by measurement of antibodies directed against b-cells. In contrast, patients converting to insulin after a latency of more than 5 years very likely have b-cell loss of non-immune nature. The true prevalence of autoimmune-based insulin deficiency amongst this patient group is not known. A previous study by our group in newly diagnosed adult-onset diabetic patients found a prevalence of anti-GAD and ICA of almost 60% in cases categorised as insulin deficient on the basis of the insulin response to a mixed meal test [16]. Type 1 diabetes represents an important subset of adult-onset diabetes mellitus. Accurate classification at diagnosis is important as it could alter management, particularly as ethical interventions for prevention or delay of b-cell destruction are developed.
References [1] M. Laasko, K. Pyrola, Age of onset and type of diabetes, Diabetes Care 8 (1985) 114 – 117. [2] R.S. Scott, L.J. Brown, Prevalence and incidence of insulin-treated diabetes mellitus in adults in Canterbury, New Zealand, Diabetic Med. 8 (1991) 1 – 8. [3] A.G. Molbak, B. Christau, B. Marner, K. Borch-Johnsen, J. Nerup, Incidence of insulin-dependent diabetes mellitus in age groups over 30 years in Denmark, Diabetic Med. 11 (1994) 650 – 655. [4] L.C. Groop, G.F. Botazzo, D. Doniach, Islet cell antibodies identify latent type 1 diabetes in patients aged 35 – 75 years at diagnosis, Diabetes 35 (1986) 237 – 241.
J. Willis et al. / Diabetes Research and Clinical Practice 42 (1998) 49–53 [5] J. Tuomilehto, P.Z. Zimmet, I.R. Mackay, et al., Antibodies to glutamic acid decarboxylase as predictors of insulin-dependent diabetes mellitus before clinical onset of disease, Lancet 343 (1994) 1383–1385. [6] L.J. Brown, R.S. Scott, A population based diabetes register — development and applications, Community Health Stud. 12 (1988) 437–441. [7] L.V. Forbes, L.J. Brown, R.S. Scott, B.A. Darlow, Immunogenetic, clinical and demographic characterisation of childhood Type 1 (insulin-dependent) diabetes in New Zealand, Diabetes Care 18 (1995) 1–8. [8] M.J. Rowley, I.R. Mackay, Q.-Y. Chen, W.J. Knowles, P.Z. Zimmet, Antibodies to glutamic acid decarboxylase discriminate major types of diabetes mellitus, Diabetes 41 (1992) 548 – 551. [9] Q.-Y. Chen, M.J. Rowley, G.C. Byrne, et al., Antibodies to glutamic acid decarboxylase in Australian children with insulin dependent diabetes mellitus and their first degree relatives, Pediatr. Res. 34 (1993) 785–790. [10] R.S. Scott, E.A. Espiner, R.A. Donald, M.J. Ellis, Free insulin, C-peptide and glucagon profiles in Type 1 diabetes, Aust. New Zealand J. Med. 10 (1980) 145–150. [11] A. Gottsater, U. Samuelsson, S. Nilsson, A. Lernmark, G. Sundkvist, Islet cell antibodies and fasting plasma
.
[12]
[13] [14]
[15]
[16]
53
C-peptide during the first 10 yr after diagnosis in patients with diabetes mellitus diagnosed in adult age, Diabetes Nutr. Metab. 5 (1992) 243 – 248. O. Hother-Nielsen, O. Faber, N. Schwatz Sorensen, H. Beck-Nielsen, Classification of newly diagnosed diabetic patients as insulin-requiring or non-insulin-requiring based on clinical and biochemical variables, Diabetes Care 11 (1988) 531 – 537. S. Madsbad, Classification of diabetes in older adults, Diabetes Care 13 (Suppl. 2) (1990) 93 – 96. T. Urakami, M. Owada, Y. Miyamoto, T. Kitagawa, H. Matsunaga, Serial changes in the prevalence of islet cell antibodies and islet cell antibody titer in children with Type 1 diabetes of abrupt or slow onset, Diabetes Care 18 (1995) 1095 – 1099. W.A. Hagopian, A.E. Karlsen, A. Gottsater, et al., Quantitative assay using recombinant human islet glutamic acid decarboxylase (GAD65) shows that 64K autoantibody positivity at onset predicts diabetes type, J. Clin. Invest. 91 (1993) 368 – 374. J.A. Willis, R.S. Scott, L.J. Brown, et al., Islet cell antibodies and antibodies against glutamic acid decarboxylase in newly diagnosed adult-onset diabetes mellitus, Diabetes Res. Clin. Pract. 33 (1996) 89 – 97.