Epidemiologic Studies of Insulin'dependent Diabetes Mellitus

E

pidemiologic Studies of Insulin'dependent Diabetes Mellitus MORTEN CHRISTY, ANDRES GREEN, BJORN CHRISTAU, HANS KROMANN, ANDJQRN NERUP

The incidence of insulin-dependent diabetes mellitus (IDDM) in Denmark in the years 1970-1976 was 13.3 per 100,000 in the age group 0—29 yr. This incidence is almost identical to that found in 1924 in Denmark in the same age group. The prevalence of insulin-consumers is 3.2 per 1,000. Comparison is made with incidence and prevalence data from other studies of Caucasian populations. DIABETES CARE 2: 127-130, MARCH-APRIL 1979.

D

enmark is well suited for epidemiologic studies due to an ethnically uniform, nonmigrating, well-educated, cooperative population; rather uniform standards of living; and an effective public health care system, readily accessible throughout the country. There is compulsory filing of hospital records; registered data about in-patients, out-patients, and patients consulting both general practitioners and specialists; and recording in the WHO code of all diagnosis made by hospital staff. Since 1970 a common computer system has been in use for recording of diagnosis. In Denmark insulin has to be prescribed by a physician. The sale of all medicine, including insulin, is by licensed pharmacies. All men in Denmark are liable to compulsory military service unless declared unfit. Since 1953, draft board registries have recorded results of an examination of all men at the age of 18-19 yr. Since IDDM patients are not accepted for military service this diagnosis will appear in the records in every known case, assuming that no IDDM patient wants to or succeeds in hiding the diagnosis. Death certificates are issued for all persons dying in Denmark and are collected by a Central Death Certificate Registry. The National Registry and a Central Person Registry make the tracing of individuals possible throughout the country. INCIDENCE OF IDDM

Using card index diagnosis registries and filed hospital records, the age-corrected incidence of IDDM in persons 0-29 yr of age during the period 1970-1976 (1970-1974 only in region 1 of the survey area) was found to be 13.3 per 100,000 per year.

The study was based on two assumptions: (1) At onset of IDDM all patients will be referred to a hospital. (2) Referral will be made to departments of either internal medicine or pediatrics. The validity of these assumptions was controlled by a mailed questionnaire to all general practitioners and specialists in private practice in the survey area, and by comparison with data in the Central Death Certificate R.egistry. Ascertainment was estimated to be 96%. The survey area comprised two separate parts of Denmark: region I includes the capital and the counties around Copenhagen inhabited by 32% of the total Danish population; region II includes two counties in West and South Jutland predominantly rural in character and inhabited by 14% of the total population. The two regions together represent 45.4% of the total population of 4,996,106 persons. Each patient's name, age at diagnosis, sex, month of diagnosis, and exact address of residence at the time of diagnosis were recorded. All patients ages 0-29 yr at diagnosis living in the survey area and requiring insulin treatment were included—a total of 792 patients. Official Danish statistical yearbooks were used as sources of data for calculation of the age-corrected incidences. Results from the survey in region I have been published previously.1 The inclusion of region II supported the conclusions made in the first report. The incidence was constant throughout the survey period (Table 1). The incidence was identical in rural and urban areas (Table 2). No clustering of cases was seen. The seasonal variation observed with a

DIABETES CARE, VOL. 2 NO. 2, MARCH-APRIL 1979

127

EP1DEMIOLOGIC STUDIES OF DIABETES / MORTEN CHRISTY AND ASSOCIATES

TABLE 1 Annual number of diagnoses of juvenile-onset insulin-dependent diabetes mellitus in two separate regions of Denmark

TABLE 3 Seasonal variation in onset of IDDM in relation to sex Onset during

Region I Copenhagen and North Zealand

Region II West and South Jutland

1970 1971 1972 1973 1974 1975 1976

97 88 98 91 100

Nl Nl

41 45 46 51 42 43 50

133 144 142 142 43 50

1970-76

474

318

792

Year

Months 2-4

Months

Females Males

84 116

Total

200

Total 138

Nl, not included. Patients ages 0-29 yr at the time of diagnosis were identified in the records of departments of pediatrics and internal medicine in two regions of Denmark inhabited by 45.4% of the total population. A prospective study of all new cases of 1DDM in region I was performed from 1 June 1975 through 31 May 1976. 79 new cases were diagnosed during this period.

22% below-average incidence during summer months was mainly due to the male patients irrespective of age at diagnosis (average 38 cases per month, range 23-53), while the number of new female cases was almost constant throughout the year (average 28 cases per month, range 22-34) (Table 3). Incidence figures related to age at diagnosis and sex are shown in Table 4. There was a remarkable incidence peak at 12-14 yr (girls preceding boys by 1-2 yr) and a minor peak at approximately 22 yr (most marked in women). The overall incidence in men exceeded the incidence in women by 26%. At present, studies are in progress to establish the age-specific incidences of IDDM in Danes aged 30 yr or more in the same survey area.

Months 8-10

Months 11-1

75 79

82 127

94 130

154

209

224

5-7

The average number of new cases per month was 38 male and 28 female subjects. The number in individual month was within the range 53-23 for male and 34-22 for female patients. Five cases included in the calculations of incidencefiguresare not included in this table since the month of diagnosis could not be identified from the records.

of us on the island of Funen inhabited by a demographic and socioeconomic 9% sample representative of the Danish population.2 In this approach, prescriptions of insulin filed during 5 months in 1973 were used to identify all consumers of insulin in the area irrespective of age. A preliminary prevalence estimate was calculated as 3.2 per 1000. This is comparable to data reported by Larsen, who found a prevalence of 3 per 1000 in Northern Zealand during 6 months of 1977.3 Hospital records and medical certificates concerning all persons primarily notified as insulin-consumers in the Funen Study are being examined to evaluate ascertainment and to obtain a description of clinical characteristics and insulin dependence. Another approach being taken4 is the use of draft board records as a possible source of prevalence figures for male IDDM patients surviving the first 20 yr of life. Data of IDDM in all Danish male citizens born in each of the years 1949-1956 and recorded in the draft board registries are being analyzed at present. Preliminary estimates of prevalence figures are in agreement with those expected from the incidence studies.

PREVALENCE OF IDDM

Two independent studies are being performed at present using almost identical methods. One is carried out by one TABLE 2 Age-corrected incidence of juvenile-onset insulin-dependent diabetes mellitus in two separate regions of Denmark (annual number per 100,000)

Age group

Region I Copenhagen and North Zealand

Region II West and South Jutland

Total

0-14 15-29 0-29

13.7 12.8 13.2

14.3 12.4 13.4

14.0 12.7 13.3

128

TABLE 4 Age-corrected incidence of juvenile-onset insulin-dependent diabetes mellitus in a survey area inhabited by 45.4% of the Danish population (per 100,000) Male

0-4 5-9 10-14 15-19 20-24 25-29 0-29

7.7

7.4

7.6

.13.2 22.7 16.8 13.9 15.0 14.8

12.9 20.1 9.3 12.0 8.9 11.7

13.0 21.4 13.1 12.9 12.1 13.3


Female

Total

Age group

EP1DEM10L0G1C STUDIES OF DIABETES / MORTEN CHRISTY AND ASSOCIATES

INCIDENCE AND PREVALENCE DATA FROM OTHER STUDIES

T

able 5 summarizes some of the most recent studies on Caucasians. The different methods applied and the different conditions under which they have been used may explain some of the discrepancies between results found in different populations. Incidence figures established by the use of filed hospital records turned out to be remarkably similar. On the other hand, this method has been applied mainly in populations that probably are very much alike genetically and culturally; i.e., Sweden, Norway, and Denmark share a common history and almost the same language and the same standards of living, and Minnesota was the state of choice for the Scandinavian immigrants to the United States. Until comparable studies have been performed in more distantly related Caucasian populations, no conclusion can be made on whether or not incidence and prevalence changes in parallel to the genetic composition of these populations.12"13

Another matter of discussion is the data suggesting a secular increase of the incidence and prevalence of IDDM.7"14 Increasing incidence figures would indicate either an increase of the frequency of the "IDDM-gene(s)" in the population or an increasing impact of diabetogenic environmental factor(s), or both. The data on diabetes mortality in Denmark presented by Heiberg15 in 1924 and census data from the Danish Government Bureau of Statistics provide an estimate of the agecorrected incidence of IDDM (with onset at 0-29 yr of age) in 1924 of 13.6 per 100,000, which is surprisingly similar to our present day figures. The incidence of IDDM with onset in childhood in the USA was 12.5 per 100,000 in 1922 according to White16 who studied mortality data. Westlund17 in Norway, using filed hospital records, found incidence figures between 3 and 16 per 100,000 for IDDM with age at onset 0-29 yr, in the years 1925-1954, with no apparent tendency towards either a decrease or an increase. However, Ustvedt7

TABLE 5 Summary of incidence and prevalence data from other Caucasian populations reported during recent years Age-corrected Incidence (/100 ,000) Reference

Population

Study period

Palumbo 19785

Minnesota

1945-69

Method of ascertainment Filed med. records

1970 Sterky 19786

Sweden

1970-75

Filed hosp. records

Patients (age in yr) DM onset before 30 yr (n = 62): 0-14 15-29

Male

Female

14.8

11.4

Total

Prevalence (/100,000) Male

Female

0.6 3.7

0.5 2.1

19.6

IDDM onset 0-15 yr (n = 359)

1975 7

Ustvedt 1977

Norway

1956-65

1.3 Filed hosp. records

DM onset (yr): 0-4 (n == 12) 5-9 (n == 39) 10-14 (n == 33) 15-19 (n == 27) 20-24 (n =: 23) 25-29 (n == 23)

Bloom 19758

U. K.

1973-74

Question, to doctors

DM onset 0-15 yr (n = 2274)

Wadsworth 19749

U. K.

1946-72

Birth cohort followed

DM onset 0-26 yr (n = 16)

Kyllo 197810

Minnesota

1975

Question, to schools

DM onset 0-15 yr (n = 191)

Lestradetl977"

France

1975

Central registry

IDDM onset (yr) (n = 5337): 0-14 yr 0-19 yr

1975

Total

6.7 16.3 13.2 11.9 9.6 11.2


2.4 14.5 11.4 7.7 7.4 5.2

7.7 3.2

2.8

3.0 2.6

3.7 0.32

129

EPIDEM10L0GIC STUDIES OF DIABETES / MORTEN CHRISTY AND ASSOCIATES

studied IDDM incidence in the same area by the same method as Westlund and in the years 1956-1965, found evidence for a slight increase in the incidence. Nevertheless, at present we find it doubtful that significant changes in the incidence of IDDM are actually taking place within the populations studied. Seasonal variation, synchronism of incidence peaks with epidemics of viral infections, the influence of age and growth spurts on incidence and variations of incidence between different locations within a larger community, and year-toyear variation have been reported by many workers since the turn of the century, and closer investigation of such epidemiologic features may well help clarify the role of environmental and genetic factors in the development of IDDM.1'6-7'16-21 From Steno Memorial Hospital, DK-2820 Gentofte, Denmark, and The University Institute of Clinical Genetics, The University Hospital, DK-5000 Odense C, Denmark. Address reprint requests to Morten Christy, Steno Memorial Hospital, DK-2820 Gentofte, Denmark. REFERENCES 1 Christau, B., Kromann, H., Ortved Andersen, O., Christy, M., Buschard, K., Arnung, K., H^jland Kristensen, I., Peitersen, J., Steinrud, J., and Nerup, J.: Incidence, seasonal, and geographical patterns of juvenile onset insulin-dependent diabetes mellitus in Denmark. Diabetologia 13: 281-284, 1977. 2 Green, A., and Hauge, M.: The prevalence of insulindependent diabetes mellitus in a ten percent sample of the Danish population. Acta Endocrinol. Suppl. 209, abstr. no. 29, 1977. 3 Larsen, D. F.: Economic problems of insulin-dependent diabetics. Ugeskr. Leg. 140: 2338-2340, 1978. 4 Green, A.: Submitted for presentation at the Internatl. Diabetes Fed. Congress, Vienna, 1979. 5 Palumbo, P. J., and Labarthe, D. R.: The incidence of diabetes mellitus in Rochester, Minnesota, 1945-1969. Adv. Metab. Dis. 9: 13-28, 1978. 6 Sterky, G., Holmgren, G., Gustavson, K. H., Larsson, Y.,

130

Lundmark, K. M., Nilsson, K. D., Samuelson, G., Thalme, B., and Wall, S.: The incidence of diabetes mellitus in Swedish children 1970-75. Acta Pediatr. Scand. 67: 139-143, 1978. 7 Ustvedt, H. J., and Olsen, E.: Incidence of diabetes mellitus in Oslo, Norway, 1956-65. Br. J. Prev. Soc. Med. 31: 251-257, 1977. 8 Bloom, A., Hayes, T. M., and Gamble, D. R.: Register of newly diagnosed diabetic children. Brit. Med. J. 3: 580— 583, 1975. 9 Wadsworth, M. E. J., and Jarrett, R. J.: Incidence of diabetes in the first 26 years of life. Lancet 2: 1172-1174, 1974. 10 Kyllo, C. J., and Muttall, F. Q.: Prevalence of diabetes mellitus in school-age children in Minnesota. Diabetes 27: 57-60, 1978. 11 Lestradet, H., and Besse, J.: Prevalence et incidence du diabete juvenile insulino-dependant en France. Diabete et Metabolisme 3: 229-234, 1977. 12 Ryder, L. P., Andersen, E., and Svejgaard, A.: An HLA map of Europe. Hum. Hered. 28: 121-200, 1978. 13 Menozzi, P., Piazza, A., and Cavalli-Sforza, L.: Synthetic maps of human gene frequencies in Europeans. Science 201: 786-92, 1978. 14 North, A. F., Gorwitz, K., and Sultz, H. A.: A secular increase in the incidence of juvenile diabetes mellitus. J. Pediatr. 91: 706-710, 1977. 15 Heiberg, K. A., and Heiberg, P.: Die Zahl der Krankheitsfalle an Diabetes Mellitus in Danemark im Jahre 1924. Acta Med. Scand. 62: 126-130, 1925. 16 White, P.: Diabetes in Childhood and Adolescence. London, Henry Kimpton, 1933, pp. 60-63. 17 Westlund, K.: Incidence of diabetes mellitus in Oslo, Norway, 1925 to 1954. Br. J. .Prev. Soc. Med. 20: 105-116, 1966. 18 Adams, S. F.: The seasonal variation in the onset of acute diabetes. Arch. Intern. Med. 37: 861-64, 1926. 19 Gundersen, E.: Is diabetes of infectious origin? J. Infect. Dis. 41: 197-202, 1927. 20 G a m b l e , D . R . , a n d Taylor, K. W . : Seasonal variation of diabetes mellitus. Br. Med. J. 2: 631-33, 1969. 21 Rolles, C. J., Rayner, P. H. W., and Mackintosh, P.: Aetiology of juvenile diabetes. Lancet 2: 230, 1975.
