.?
ORiGlNAL ARTICLES
Familial Aggregation in Type 1 (Insulindependent) Diabetes Mellitus: a Study from South India A. Ramachandran, C. Snehalatha, 1. Premila, V. Mohan, M. Viswanathan Diabetes Research Centre and M. V . Hospiiaf for Diabetes, Royapuram, Madras, lndra
An analysis of 617 Type 1 (insulin-dependent) diabetic patients from 587 families was made. In 33 families (5.6 %) there was more than one Type 1 diabetic patient. In 98 families (16.7 %) positive family history of Type 2 (non-insulin-dependent) diabetes was present. There was a linear correlation between the age at diagnosis of the proband and that of the sibling ( r = 0.73, p < 0.001) in the Type 1 multiplex families. The risk of Type 1 diabetes in relatives was found to be 2.7 % by Li Mantel estimate, which i s lower than that reported among Europeans. The cumulative risk was higher ( p < 0.001) in the relatives of Type 1 diabetic patients with age at diagnosis I10 years. KEY WORDS
Type 1 (insulin-dependent) diabetes Familial aggregation Family history of diabetes Risk of diabetes India
Introduction The limited data on the risk of diabetes in family members of Type 1 diabetic patients have come from studies on European population^.'-^ There are no reports on this from outside Europe and the USA. Furthermore nothing is known about the risk of Type 1 diabetes in Indian populations. In this paper we report on the familial aggregation and the risk of diabetes among the relatives of Type 1 diabetic patients at a large diabetes centre in Southern India.
Patients and Methods All Type 1 diabetic patients attending the Diabetes Research Centre and M.V. Hospital for Diabetes, Madras, are recorded in a special register. Six-hundred and seventeen subjects who visited the clinic as new and review cases were registered during the period from January 1984 to January 1988. The majority of the subjects came from the four southern states of India namely Tamil Nadu, Andhra Pradesh, Karnataka, and Kerala. Diagnosis as Type 1 diabetes was made according to the classification of the WHO expert committee.6 These patients had acute onset of symptoms, proneness to ketosis, and absolute dependence on insulin from the time of diagnosis. The time when the first diagnostic test was performed was recorded as the month of diagnosis. A detailed family history was taken in respect of first degree relatives (father, mother, siblings, and offspring). A questionnaire was used to elicit all relevant details of diabetes including Correspondence to: Dr A. Ramachandran, Diabetes Research Centre, 5, Main Road, Royapuram, Madras 600 01 3, India.
876
0742-3071/90/100876-04 $05.00 & Sons, Ltd.
0 1990 by john Wiley
the age at diagnosis, the present age, type of treatment, and clinical status of the parents and siblings. This included the details of any deceased diabetic relatives. The accuracy of this was confirmed by questioning at least one more member of the family wherever possible. Five probands and 2 of their siblings with Type 1 diabetes died after registration. Diagnosis as Type 2 diabetes was based on adequate control of diabetes for a minimum period of 5 years without insulin. Some of them had required insulin for short periods, during conditions of stress, but had never used insulin alone continuously. Diagnosis was verified from the details of the physicians’ notes and the treatment. The 61 7 Type 1 diabetic patients recorded in this study belonged to 587 families and all of them were seen at least once at our centre. The first patient seen at the centre from a multiplex family was considered as the proband in this study. The rnale:female ratio was 1.5:l.
Statistical Methods Student‘s t-test, Spearman’s linear correlation, and test of proportions were used. To calculate the risk to relatives, the proband exclusion procedure as described by Wagener et al.4 and the Li Mantel test’ were used (Appendix).
Results Family History of Type 7 Diabetes The 617 patients belonged to 587 families. Thus 33 (5.6 %) families had more than one Type 1 diabetic Accepted 2 July 1990 DIABETIC MEDICINE, 1990; 7: 876-879
patient, of whom 30 were first degree relatives (Type 1 multiplex families). In 23 multiplex families, there were 26 siblings with diabetes (multiplex sibling families). The details of the family history are shown in Table 1. A positive family history of Type 2 diabetes was obtained in 98 families (16.7 %). This included diabetes in a parent in 82 families.
35 55 40 30
20
There was a linear correlation between the age at diagnosis of the proband and that of the sibling (r = 0.73, p < 0.001) as shown in Figure 1. Analysis of the age at diagnosis in the multiplex families according to the family size is shown in Table 2. A majority of the families with multiplex siblings (20/26) had large families (> 3 offspring). The mean age at diagnosis of the diabetic patients in the single case families was higher in large families ( p < 0.001). This difference was not seen in the multiplex families.
15
The risk of developing Type 1 diabetes in relatives of the probands calculated by the proband exclusion procedure was 1.5 % and that by the Li Mantel estimate was 2.7 % (Appendix 1 ). The cumulative risk of diabetes in the first degree relatives of the probands (according to the age at
Table 1. Family history of diabetes in the Type 1 diabetic study subjects Number of fami Iies Type I diabetes First degree relatives One sib Type 1 One parent Type 2 + one or more sib Type 1 One parent Type 1 Offspring Type 1 Both parents Type 2 + one or more sib Type 1 Second degree relatives Other Type 1 Type 2 diabetes First degree relatives One parent Type 2 One parent Type 2
+
one or more sib Type 2 One or more sib Type 2 Both parents Type 2 Both parents Type 2 one or more sib Type 2
+
No family history All families
FAMILIAL AGGREGATION IN TYPE 1 DIABETES
0
0
LY I
0O
10 5 J
0
J
5 10 15 20 25 30 35 40 PRC8AND-AGE AT DIAGNOSIS (YEARS)
Figure 1. Correlation between age at diagnosis of the proband and age at diagnosis of the sibling in multiplex families. Regression analysis shows a good correlation ( r = 0.73, p < 0.001) between the age of onset of the probands and their siblings; y = 4.4 + 0 . 7 ~
diagnosis of the proband) is shown in Table 3. The cumulative risk was significantly higher ( p < 0.001) in the relatives of Type 1 probands with age at diagnosis 10 years or below.
Prevalence of Type 2 Diabetes in the Type 7 Diabetic Families
%
33
5.6
16
2.7
6 5 2
0.9 0.3
1
0.2
3
0.5
98
16.7
65
11.0
17
9 4
2.9 1.5 0.6
3
0.5
456 587
0
25
Multiplex Families
Risk of Type 1 Diabetes in Relatives
1
1
.o
The age of all members could be ascertained in 1963 first degree relatives belonging to 307 families. Table 4 shows the age-specific prevalence rate of Type 2 diabetes in the relatives compared with the age-specific prevalence in the general population. There was no significant difference in the prevalence of Type 2 diabetes in these families (Test of Proportions).
Discussion This is the first report on a family study in Type 1 diabetes from the Indian subcontinent. Out of a total of 587 families with Type 1 diabetes, 30 families (5 %) had more than one Type 1 diabetic patient in first degree relatives (Type 1 multiplex). Another 3 had family history of Type 1 diabetes in other relatives, and thus 33 families (5.6 %) had family history of Type 1 diabetes. This figure is much lower than those reported from European populations. Dalquist et a / . 8 reported that 8.5 % of Swedish diabetic children had a diabetic parent. In the USA Chern et a/.5 reported that 16 % of cases of Type 1 diabetes belonged to multiplex families. The possibilities for a false low estimate include death of relatives with Type 1 diabetes before registration and death of siblings 877
Dm
ORlGiNAL ARTICLES Table 2. Age at diagnosis (years) of multiplex families with Type 1 diabete5 according to family size Single case families
Families with 1-3 offspring Families with > 3 offspring
16 k 12 (257) 25 t 13 (297)”
Families with multiplex siblings First case Other cases
13 i 8 (3) 16 10 (20) +_
9 2 5 (3) 14 i 8 (23)
Mean 2 SD (number). “ p < 0.001 compared with tamilies with 1-3 offspring Table 3. Cumulative ri5k of Type 1 diabetes among the siblings Total number
Number at risk
Sibs of probands with age at onset Age of sibs (yr)
10 11-20 21-30
8 16 4
5
10 11-20 2 1-30 31-40 41 -50
1 13 13 15 3
”p
< 0.001 with
s 10 years
< 20 20-24 25-34 35-44 45-54 55-64 =’ 65
45 44 31 18 3
All ages
First degree relatives n
Z Type 2 diabetes Prevalence n % in general valuesb populationa %
339 175 399 385 31 1 213 141 1963
1 1 3 23 44 43 33 148
0 0 2.1 5.9 19.0 29.4 28.6
-
-
0.48 0.96 0.06 0.92 0.65 0.24 -
”See Ramachandran et a/. ‘‘I bNot significant.
at an early age before reaching the age of risk of diabetes. The siblings who died of Type 1 diabetes have been included in this analysis. We have also taken the details of causes of death of children prior to the registration in this study and none of them had a record of having Type 1 diabetes. The second phenomenon i s true for all populations. Thus, the low prevalence of Type 1 diabetes
878
10.7 30.7 55.7”
0.02 0.1 1 0.19
2.2 13.6 33.0 33.0 33.0
-
respect to sibs of probsnd with age at onwi
0.3 0.6 0.8 6.0 14.1 20.2 23.4
0.11 0.20 0.25
> 10 years
Table 4. Type 2 diabetes in 307 families of Type 1 diabetic patients Age group (years)
Cumulative Cumulative risk (%) proportion developing Tvpe 1 diabetes
28 20 4
Sibs of probands with age at onset Age of sibs (yr) 5
Number with d I abetes
>
1 0 tiears
among the relatives of our patients is probably not related to the above causes.
Employing the proband exclusion and Li Mantel estimates, the risk of diabetes in the relatives of those with Type 1 diabetes was found to be 1.5 Yo and 2.7 %, respectively. These figures are significantly lower than those in Europeans4 (3.2 % and 6 Yo, respectively). This is, however, similar to the estimate of 1.6 % from Canada by Simpson.‘ The reason for this low risk of diabetes in the relatives of our patients remains unclear. It may therefore be a reflection of the genetic and ethnic variations present in the manifestation of diabetes. The siblings of Type 1 diabetic probands with age at diagnosis 5 10 years had significantly higher risk of developing the disease than the siblings of probands who developed diabetes after the age of 10 years. The risk increased with increasing age of the sibling, especially in the former group. This observation supports the contention that the risk of diabetes is greater among the siblings of younger Type 1 diabetic patients. This is in agreement with the reports of Gamble,’ Chern et a/.,‘j and Tillil et a/.2 We had chosen the age groups of 5 10 years and above to compare the observations with those already reported in the above studies. It is interesting that there is a linear correlation between the age of diagnosis of siblings in multiplex families, as shown in A. R A M A C H A N D R A N E l A l .
DT17 Figure 1. This indicates that the genetic determinants decide the age at which the disorder manifests itself in a family. This may not be true in all populations. For example, Gamble' noted that age at onset of the disease in sibling pairs were independently determined in the children in the UK. The mean age of diagnosis of the second case in a multiplex family was not different from that of the first case. Wagener et a/.4 in the Pittsburgh study of Type 1 diabetes reported similar findings. The age at diagnosis of Type 1 diabetes in single case families with a larger number of offspring was significantly higher than in the smaller families ( p < 0.001). Moreover, a majority of the multiplex families had large family size. The hypothesis put forward by Wagener et a/.4 that children from larger families have lived through longer periods of risk and therefore develop diabetes at an older age is supported by this. However, contrary to the data presented in that report, the age at diagnosis of our multiplex cases was less than that in the single case families. The prevalence of Type 2 diabetes in families of the Type 1 probands is much higher in our study (16.7 %) than in reports in Eur~peans.~," According to Wagener et a/.4, the increase of Type 2 diabetes among the relatives of Type 1 diabetic patients may be an artifact due to increased awareness about diabetes. We found no significant difference in the age-specific prevalence of diabetes in the relatives compared with the general population. Thus while the prevalence of Type 2 diabetes among the relatives was higher than that seen in Europeans, it i s not different from the prevalence rate in our population as a whole.'o Acknowledgements We thank R.E. LaPorte and J.S. Dorman of the University of Pittsburgh, USA, for their suggestions in the preparation of the manuscript.
ORIGINAL ARTICLES 2.
1980; 19: 341-344. Tillil ti, Knuppel S, Kobberling J. Early onset of type I
(insulin dependent) diabetes is associated with an increased risk for Type I diabetes in siblings (Abstract). Diabetologia 1989; 23: 5 1 OA. 3. Warram JH, Krolewski AS, Gottilieb MS, Kahn CR. Ditferences in risk of insulin dependent diabetes in offspring of diabetic mothers and diabetic fathers. N Engl 1 Med 1984; 311: 49-52. 4. Wagener DK, Sacks JM, LaPorte RE, MacGregor JM.The Pittsburgh study of insulin dependent diabete3 mellitus: risk of diabetes among relatives of IDDM. Diabetes 1982;
31: 136-144. Chern MM, Anderson VE, Barbosa 1. Empirical risk for insulin-dependent diabetes (IDD) in sibs: further definitions of genetic heterogeneity. Diabetes 1982; 31: 1 1 1 5- 1 1 1 8. 6. W H O Study Group. Diabetes Mellitus. Technical Report Series 727. Geneva: WHO, 1985. 7. Li CC, Mantel N. A simple method of estimating the segregation ratio under complete ascertainment. Am Horn Cmet 1968; 20: 61-81. 8. Dalquist G, Bloom L, Tuvemo T, Nystrom L, Sandstrom A, Wall S. The Swedish childhood diabetes study: results from a nine year case register and a one year case referent study indicating that type I (insulin-dependent) diabetes mellitus is associated with both type / I (non-insulin dependent) diabetes rnellitus and autoimmune disorders. Diabetologia 1989; 32: 2-6. 9. Simpson NE. Heritabilities of vulnerability to diabetes when sex and age at onset are considered. Ann Hum Genet 1969; 32: 283-286. 10. Ramachandran A, Jali MV, Mohan V, Snehalatha C, Viswanathan M. High prevalence of diabetes in an urban population in South India. Br Med / 1988; 297: 587-590.
5.
Appendix: Calculation of Risk in Relatives 1. Proband exclusion
2. Li Mantel
diabetes affecting two or more siblings. Diabetologia
FAMILIAL AGGREGATION IN TYPE 1 DIABETES
/
~~
=
1.5 %
7- N
241 3 -590
R-/
617-567 = 2.7 % 2413-567
r=--=--
J-/
References 1 . Gamble DR. An epidemiological study of childhood
R-N - 61 7-590 r=---
= total number of sibs; R = affected number of sibs; = number of single proband families; N = total number of
families.
879
ORiGlNAL ARTICLES
Familial Aggregation in Type 1 (Insulindependent) Diabetes Mellitus: a Study from South India A. Ramachandran, C. Snehalatha, 1. Premila, V. Mohan, M. Viswanathan Diabetes Research Centre and M. V . Hospiiaf for Diabetes, Royapuram, Madras, lndra
An analysis of 617 Type 1 (insulin-dependent) diabetic patients from 587 families was made. In 33 families (5.6 %) there was more than one Type 1 diabetic patient. In 98 families (16.7 %) positive family history of Type 2 (non-insulin-dependent) diabetes was present. There was a linear correlation between the age at diagnosis of the proband and that of the sibling ( r = 0.73, p < 0.001) in the Type 1 multiplex families. The risk of Type 1 diabetes in relatives was found to be 2.7 % by Li Mantel estimate, which i s lower than that reported among Europeans. The cumulative risk was higher ( p < 0.001) in the relatives of Type 1 diabetic patients with age at diagnosis I10 years. KEY WORDS
Type 1 (insulin-dependent) diabetes Familial aggregation Family history of diabetes Risk of diabetes India
Introduction The limited data on the risk of diabetes in family members of Type 1 diabetic patients have come from studies on European population^.'-^ There are no reports on this from outside Europe and the USA. Furthermore nothing is known about the risk of Type 1 diabetes in Indian populations. In this paper we report on the familial aggregation and the risk of diabetes among the relatives of Type 1 diabetic patients at a large diabetes centre in Southern India.
Patients and Methods All Type 1 diabetic patients attending the Diabetes Research Centre and M.V. Hospital for Diabetes, Madras, are recorded in a special register. Six-hundred and seventeen subjects who visited the clinic as new and review cases were registered during the period from January 1984 to January 1988. The majority of the subjects came from the four southern states of India namely Tamil Nadu, Andhra Pradesh, Karnataka, and Kerala. Diagnosis as Type 1 diabetes was made according to the classification of the WHO expert committee.6 These patients had acute onset of symptoms, proneness to ketosis, and absolute dependence on insulin from the time of diagnosis. The time when the first diagnostic test was performed was recorded as the month of diagnosis. A detailed family history was taken in respect of first degree relatives (father, mother, siblings, and offspring). A questionnaire was used to elicit all relevant details of diabetes including Correspondence to: Dr A. Ramachandran, Diabetes Research Centre, 5, Main Road, Royapuram, Madras 600 01 3, India.
876
0742-3071/90/100876-04 $05.00 & Sons, Ltd.
0 1990 by john Wiley
the age at diagnosis, the present age, type of treatment, and clinical status of the parents and siblings. This included the details of any deceased diabetic relatives. The accuracy of this was confirmed by questioning at least one more member of the family wherever possible. Five probands and 2 of their siblings with Type 1 diabetes died after registration. Diagnosis as Type 2 diabetes was based on adequate control of diabetes for a minimum period of 5 years without insulin. Some of them had required insulin for short periods, during conditions of stress, but had never used insulin alone continuously. Diagnosis was verified from the details of the physicians’ notes and the treatment. The 61 7 Type 1 diabetic patients recorded in this study belonged to 587 families and all of them were seen at least once at our centre. The first patient seen at the centre from a multiplex family was considered as the proband in this study. The rnale:female ratio was 1.5:l.
Statistical Methods Student‘s t-test, Spearman’s linear correlation, and test of proportions were used. To calculate the risk to relatives, the proband exclusion procedure as described by Wagener et al.4 and the Li Mantel test’ were used (Appendix).
Results Family History of Type 7 Diabetes The 617 patients belonged to 587 families. Thus 33 (5.6 %) families had more than one Type 1 diabetic Accepted 2 July 1990 DIABETIC MEDICINE, 1990; 7: 876-879
patient, of whom 30 were first degree relatives (Type 1 multiplex families). In 23 multiplex families, there were 26 siblings with diabetes (multiplex sibling families). The details of the family history are shown in Table 1. A positive family history of Type 2 diabetes was obtained in 98 families (16.7 %). This included diabetes in a parent in 82 families.
35 55 40 30
20
There was a linear correlation between the age at diagnosis of the proband and that of the sibling (r = 0.73, p < 0.001) as shown in Figure 1. Analysis of the age at diagnosis in the multiplex families according to the family size is shown in Table 2. A majority of the families with multiplex siblings (20/26) had large families (> 3 offspring). The mean age at diagnosis of the diabetic patients in the single case families was higher in large families ( p < 0.001). This difference was not seen in the multiplex families.
15
The risk of developing Type 1 diabetes in relatives of the probands calculated by the proband exclusion procedure was 1.5 % and that by the Li Mantel estimate was 2.7 % (Appendix 1 ). The cumulative risk of diabetes in the first degree relatives of the probands (according to the age at
Table 1. Family history of diabetes in the Type 1 diabetic study subjects Number of fami Iies Type I diabetes First degree relatives One sib Type 1 One parent Type 2 + one or more sib Type 1 One parent Type 1 Offspring Type 1 Both parents Type 2 + one or more sib Type 1 Second degree relatives Other Type 1 Type 2 diabetes First degree relatives One parent Type 2 One parent Type 2
+
one or more sib Type 2 One or more sib Type 2 Both parents Type 2 Both parents Type 2 one or more sib Type 2
+
No family history All families
FAMILIAL AGGREGATION IN TYPE 1 DIABETES
0
0
LY I
0O
10 5 J
0
J
5 10 15 20 25 30 35 40 PRC8AND-AGE AT DIAGNOSIS (YEARS)
Figure 1. Correlation between age at diagnosis of the proband and age at diagnosis of the sibling in multiplex families. Regression analysis shows a good correlation ( r = 0.73, p < 0.001) between the age of onset of the probands and their siblings; y = 4.4 + 0 . 7 ~
diagnosis of the proband) is shown in Table 3. The cumulative risk was significantly higher ( p < 0.001) in the relatives of Type 1 probands with age at diagnosis 10 years or below.
Prevalence of Type 2 Diabetes in the Type 7 Diabetic Families
%
33
5.6
16
2.7
6 5 2
0.9 0.3
1
0.2
3
0.5
98
16.7
65
11.0
17
9 4
2.9 1.5 0.6
3
0.5
456 587
0
25
Multiplex Families
Risk of Type 1 Diabetes in Relatives
1
1
.o
The age of all members could be ascertained in 1963 first degree relatives belonging to 307 families. Table 4 shows the age-specific prevalence rate of Type 2 diabetes in the relatives compared with the age-specific prevalence in the general population. There was no significant difference in the prevalence of Type 2 diabetes in these families (Test of Proportions).
Discussion This is the first report on a family study in Type 1 diabetes from the Indian subcontinent. Out of a total of 587 families with Type 1 diabetes, 30 families (5 %) had more than one Type 1 diabetic patient in first degree relatives (Type 1 multiplex). Another 3 had family history of Type 1 diabetes in other relatives, and thus 33 families (5.6 %) had family history of Type 1 diabetes. This figure is much lower than those reported from European populations. Dalquist et a / . 8 reported that 8.5 % of Swedish diabetic children had a diabetic parent. In the USA Chern et a/.5 reported that 16 % of cases of Type 1 diabetes belonged to multiplex families. The possibilities for a false low estimate include death of relatives with Type 1 diabetes before registration and death of siblings 877
Dm
ORlGiNAL ARTICLES Table 2. Age at diagnosis (years) of multiplex families with Type 1 diabete5 according to family size Single case families
Families with 1-3 offspring Families with > 3 offspring
16 k 12 (257) 25 t 13 (297)”
Families with multiplex siblings First case Other cases
13 i 8 (3) 16 10 (20) +_
9 2 5 (3) 14 i 8 (23)
Mean 2 SD (number). “ p < 0.001 compared with tamilies with 1-3 offspring Table 3. Cumulative ri5k of Type 1 diabetes among the siblings Total number
Number at risk
Sibs of probands with age at onset Age of sibs (yr)
10 11-20 21-30
8 16 4
5
10 11-20 2 1-30 31-40 41 -50
1 13 13 15 3
”p
< 0.001 with
s 10 years
< 20 20-24 25-34 35-44 45-54 55-64 =’ 65
45 44 31 18 3
All ages
First degree relatives n
Z Type 2 diabetes Prevalence n % in general valuesb populationa %
339 175 399 385 31 1 213 141 1963
1 1 3 23 44 43 33 148
0 0 2.1 5.9 19.0 29.4 28.6
-
-
0.48 0.96 0.06 0.92 0.65 0.24 -
”See Ramachandran et a/. ‘‘I bNot significant.
at an early age before reaching the age of risk of diabetes. The siblings who died of Type 1 diabetes have been included in this analysis. We have also taken the details of causes of death of children prior to the registration in this study and none of them had a record of having Type 1 diabetes. The second phenomenon i s true for all populations. Thus, the low prevalence of Type 1 diabetes
878
10.7 30.7 55.7”
0.02 0.1 1 0.19
2.2 13.6 33.0 33.0 33.0
-
respect to sibs of probsnd with age at onwi
0.3 0.6 0.8 6.0 14.1 20.2 23.4
0.11 0.20 0.25
> 10 years
Table 4. Type 2 diabetes in 307 families of Type 1 diabetic patients Age group (years)
Cumulative Cumulative risk (%) proportion developing Tvpe 1 diabetes
28 20 4
Sibs of probands with age at onset Age of sibs (yr) 5
Number with d I abetes
>
1 0 tiears
among the relatives of our patients is probably not related to the above causes.
Employing the proband exclusion and Li Mantel estimates, the risk of diabetes in the relatives of those with Type 1 diabetes was found to be 1.5 Yo and 2.7 %, respectively. These figures are significantly lower than those in Europeans4 (3.2 % and 6 Yo, respectively). This is, however, similar to the estimate of 1.6 % from Canada by Simpson.‘ The reason for this low risk of diabetes in the relatives of our patients remains unclear. It may therefore be a reflection of the genetic and ethnic variations present in the manifestation of diabetes. The siblings of Type 1 diabetic probands with age at diagnosis 5 10 years had significantly higher risk of developing the disease than the siblings of probands who developed diabetes after the age of 10 years. The risk increased with increasing age of the sibling, especially in the former group. This observation supports the contention that the risk of diabetes is greater among the siblings of younger Type 1 diabetic patients. This is in agreement with the reports of Gamble,’ Chern et a/.,‘j and Tillil et a/.2 We had chosen the age groups of 5 10 years and above to compare the observations with those already reported in the above studies. It is interesting that there is a linear correlation between the age of diagnosis of siblings in multiplex families, as shown in A. R A M A C H A N D R A N E l A l .
DT17 Figure 1. This indicates that the genetic determinants decide the age at which the disorder manifests itself in a family. This may not be true in all populations. For example, Gamble' noted that age at onset of the disease in sibling pairs were independently determined in the children in the UK. The mean age of diagnosis of the second case in a multiplex family was not different from that of the first case. Wagener et a/.4 in the Pittsburgh study of Type 1 diabetes reported similar findings. The age at diagnosis of Type 1 diabetes in single case families with a larger number of offspring was significantly higher than in the smaller families ( p < 0.001). Moreover, a majority of the multiplex families had large family size. The hypothesis put forward by Wagener et a/.4 that children from larger families have lived through longer periods of risk and therefore develop diabetes at an older age is supported by this. However, contrary to the data presented in that report, the age at diagnosis of our multiplex cases was less than that in the single case families. The prevalence of Type 2 diabetes in families of the Type 1 probands is much higher in our study (16.7 %) than in reports in Eur~peans.~," According to Wagener et a/.4, the increase of Type 2 diabetes among the relatives of Type 1 diabetic patients may be an artifact due to increased awareness about diabetes. We found no significant difference in the age-specific prevalence of diabetes in the relatives compared with the general population. Thus while the prevalence of Type 2 diabetes among the relatives was higher than that seen in Europeans, it i s not different from the prevalence rate in our population as a whole.'o Acknowledgements We thank R.E. LaPorte and J.S. Dorman of the University of Pittsburgh, USA, for their suggestions in the preparation of the manuscript.
ORIGINAL ARTICLES 2.
1980; 19: 341-344. Tillil ti, Knuppel S, Kobberling J. Early onset of type I
(insulin dependent) diabetes is associated with an increased risk for Type I diabetes in siblings (Abstract). Diabetologia 1989; 23: 5 1 OA. 3. Warram JH, Krolewski AS, Gottilieb MS, Kahn CR. Ditferences in risk of insulin dependent diabetes in offspring of diabetic mothers and diabetic fathers. N Engl 1 Med 1984; 311: 49-52. 4. Wagener DK, Sacks JM, LaPorte RE, MacGregor JM.The Pittsburgh study of insulin dependent diabete3 mellitus: risk of diabetes among relatives of IDDM. Diabetes 1982;
31: 136-144. Chern MM, Anderson VE, Barbosa 1. Empirical risk for insulin-dependent diabetes (IDD) in sibs: further definitions of genetic heterogeneity. Diabetes 1982; 31: 1 1 1 5- 1 1 1 8. 6. W H O Study Group. Diabetes Mellitus. Technical Report Series 727. Geneva: WHO, 1985. 7. Li CC, Mantel N. A simple method of estimating the segregation ratio under complete ascertainment. Am Horn Cmet 1968; 20: 61-81. 8. Dalquist G, Bloom L, Tuvemo T, Nystrom L, Sandstrom A, Wall S. The Swedish childhood diabetes study: results from a nine year case register and a one year case referent study indicating that type I (insulin-dependent) diabetes mellitus is associated with both type / I (non-insulin dependent) diabetes rnellitus and autoimmune disorders. Diabetologia 1989; 32: 2-6. 9. Simpson NE. Heritabilities of vulnerability to diabetes when sex and age at onset are considered. Ann Hum Genet 1969; 32: 283-286. 10. Ramachandran A, Jali MV, Mohan V, Snehalatha C, Viswanathan M. High prevalence of diabetes in an urban population in South India. Br Med / 1988; 297: 587-590.
5.
Appendix: Calculation of Risk in Relatives 1. Proband exclusion
2. Li Mantel
diabetes affecting two or more siblings. Diabetologia
FAMILIAL AGGREGATION IN TYPE 1 DIABETES
/
~~
=
1.5 %
7- N
241 3 -590
R-/
617-567 = 2.7 % 2413-567
r=--=--
J-/
References 1 . Gamble DR. An epidemiological study of childhood
R-N - 61 7-590 r=---
= total number of sibs; R = affected number of sibs; = number of single proband families; N = total number of
families.
879
