ArticlePDF Available

Worldwide Increase in Incidence of Type I Diabetes – The Analysis of the Data on Published Incidence Trends


Abstract and Figures

Several reports on the incidence of Type I (insulin-dependent) diabetes mellitus have suggested that the incidence is increasing. The aim of this study was to find out whether the incidence is increasing globally or restricted to a selected populations only and to estimate the magnitude of the change in incidence. During 1960 to 1996 37 studies in 27 countries were carried out. To fulfil the inclusion criteria the study periods ranged from 8-32 years. The temporal trend was fitted by linear regression, with the logarithm of the age-standardized incidence as the dependent variable and the calendar year as the independent variable. Then, the regression coefficient (x 100%) is approximately the average relative increase in incidence per year (as percentage). Results from the pooled data from all 37 populations showed that the overall increase in incidence was 3.0% per year (95% CI 2.6; 3.3, p = 0.0001). The statistically significant increase was found in 24 of 37 populations including all high incidence (> 14.6 per 100000 a year) populations. The relative increase was, however, steeper in the populations with a lower incidence. The correlation between logarithm of the incidence and the increase in incidence was r = -0.56, p = 0.0004. The incidence of Type I diabetes is increasing worldwide both in low and high incidence populations. By the year 2010 the incidence will be 50 per 100000 a year in Finland and also in many other populations it will exceed 30 per 100000 a year.
Content may be subject to copyright.
The aetiology and natural history of Type I (insulin-
dependent) diabetes mellitus are still not known but
both genetics and environmental factors contribute
to the development of the disease [1±3]. Although
HLA genetics have a major role in the aetiology of
Type I diabetes, other genes also contribute to the ge-
netic effect, but the mode of inheritance of the dis-
ease is not clear [4]. The genetic effect contributes
70±75% of the susceptibility to Type I diabetes [5,
6]. Environmental factors possibly initiate or trigger
the process which leads to the destruction of the
beta cells and the onset of diabetes [3, 7, 8].
In the late 1970s epidemiological reports of dia-
betic children for the first time showed a wide geo-
graphical variation in the incidence of Type I diabe-
tes. During the 1960s to the early 1980s the data on
incidence of Type I diabetes were available for a few
populations only, mostly from regions with a high or
intermediate risk for this disease. A large number of
registries had been established since the mid 1980s
worldwide. The lack of standardized data made it dif-
ficult to determine the true magnitude of the world-
wide variation in incidence or time trends [9]. The Di-
Diabetologia (1999) 42: 1395±1403
Worldwide increase in incidence of Type I diabetes ±
the analysis of the data on published incidence trends
P. Onkamo, S. Väänänen, M. Karvonen, J. Tuomilehto
Diabetes and Genetic Epidemiology Unit, Department of Epidemiology and Health Promotion, National Public Health Institute,
Helsinki, Finland
Ó Springer-Verlag 1999
Aims/hypothesis. Several reports on the incidence of
Type I (insulin-dependent) diabetes mellitus have
suggested that the incidence is increasing. The aim
of this study was to find out whether the incidence is
increasing globally or restricted to a selected popula-
tions only and to estimate the magnitude of the
change in incidence.
Methods. During 1960 to 1996 37 studies in 27 coun-
tries were carried out. To fulfil the inclusion criteria
the study periods ranged from 8±32 years. The tem-
poral trend was fitted by linear regression, with the
logarithm of the age-standardized incidence as the
dependent variable and the calendar year as the inde-
pendent variable. Then, the regression coefficient
( ´ 100%) is approximately the average relative in-
crease in incidence per year (as percentage).
Results. Results from the pooled data from all 37 pop-
ulations showed that the overall increase in incidence
was 3.0% per year (95% CI 2.6; 3.3, p = 0.0001). The
statistically significant increase was found in 24 of 37
populations including all high incidence ( > 14.6 per
100 000 a year) populations. The relative increase
was, however, steeper in the populations with a lower
incidence. The correlation between logarithm of the
incidence and the increase in incidence was
r = ±0.56, p = 0.0004.
Conclusion/interpretation. The incidence of Type I di-
abetes is increasing worldwide both in low and high
incidence populations. By the year 2010 the incidence
will be 50 per 100 000 a year in Finland and also in
many other populations it will exceed 30 per 100 000
a year. [Diabetologia (1999) 42: 1395±1403]
Keywords Epidemiology, geographical variation, in-
cidence, increase, modelling, prediction, trend, Type
I diabetes.
Received: 3 May 1999 and in final revised form: 17 August
Corresponding author: J. Tuomilehto, Diabetes and Genetic
Epidemiology Unit, Department of Epidemiology and Health
Promotion, National Public Health Institute, Mannerheimin-
tie 166, FIN-00300 Helsinki, Finland
abetes Epidemiology Research International Group
(DERI) started the collection of the aggregate data
on incidence of Type I diabetes in the late 1970s and
the early 1980s [10]. The efforts of the DERI group
led to an increase in the number of registries on dia-
betic children and to the establishment of the World
Health Organization Project of Childhood Diabetes
(DIAbetes MONdiale) in 1990 [11]. The collabora-
tive research project EURODIAB ACE was estab-
lished also in the late 1980s [12] to gather information
of Type I diabetes in Europe. The latest reviews on
the incidence of Type I diabetes among populations
have indicated that differences in the incidence are
60-fold between the highest and the lowest rates [10,
12±14]. The highest incidence is found in Caucasoid
populations, particularly in northern Europe, and
the lowest rates are found in Asia and South America
[13, 14].
Thus far only one trend analysis of the incidence of
Type I diabetes comparing simultaneously several
but yet a limited number of populations has been car-
ried out by the DERI group [15]. The standardized
procedures agreed upon for the incidence data col-
lected around the world now permit a comparative
assessment of temporal trends among several popula-
tions. We estimated the temporal trends in the inci-
dence of Type I diabetes from incidence data collect-
ed through a systematic literature review. A statistical
analysis of the data was done in order to find out
whether the incidence is increasing globally. Another
objective was to evaluate quantitatively the extent to
which the change in incidence of Type I diabetes dif-
fers among populations.
Materials and methods
Literature search. The literature was searched using MED-
LINE, direct examination of reference lists of the articles,
hand searches of selected journals and published conference
abstracts. By the closing date, 28th February 1999, more than
160 original publications reporting time series of the incidence
of Type I diabetes were found.
Inclusion criteria. The publications were further evaluated
with the strict inclusion criteria in order to choose appropri-
ate studies for the quantitative analysis. The inclusion criteria
were: 1) the study period was 8 years or more, 2) the inci-
dence rates were presented for each year separately, 3) the
number of cases per year was five or more, 4) in the papers
in which the age standardization had been reported, the inci-
dences had been estimated with age standardization accord-
ing to the world population and, 5) Type I diabetes was diag-
nosed according to the WHO definition. Studies comparing
incidence rates estimated with different methods during dif-
ferent periods were excluded. No requirement for a minimum
case-ascertainment level of the data source was made as reli-
able case ascertainment estimation had usually not been
done until the 1990 s. The articles were evaluated by two inde-
pendent reviewers according to the above-mentioned inclu-
sion criteria.
Description of the data. Incidence data were obtained either
from the tables or from the figures in the published articles.
In approximately half of the original articles the annual inci-
dence rates were presented in the tables and in another half in
the figures only. The numbers derived from the figures were
reconfirmed by the second reviewer. For Montreal (Canada),
Allegheny County (USA), Scotland, Auckland (New Zea-
land), Prince Edward Island (Canada), Leicestershire (UK)
and Wielkopolska (Poland) we used the original database of
the DERI Study [15].
Altogether 37 studies from 27 countries met the inclusion
criteria and were included in the analysis (Table 1). The regis-
tration of diabetic children was prospective in most of the
studies. In 30 studies the age of children ranged from 0 to
14 years and in 7 studies from 0 to 15, 16, 17 and 19 years.
The time period of the studies ranged from 8 to 32 years.
The average length of the study period was 14.9 years (medi-
an 14 years). The estimates of the degree of case-ascertain-
ment were high, ranging from 85 to 100%. The degree of as-
certainment remained unspecified only in five studies. The
studies included in the analysis were from the period 1960 to
Statistical methods. The incidence of Type I diabetes for our
analysis was taken from the individual studies as it was report-
ed in these publications. The incidence for the data obtained
from the DERI Study [15] was calculated per 100 000 people
a year. Age standardization of the rates was done using 5-year
intervals with the proportions 33/100, 33/100 and 34/100 (for
the age groups 0±4 years, 5±9 years, and 10±14 years respec-
tively) as the standard according to the approach by the
DERI Study Group [16], which is the same as the world popu-
lation standard.
The temporal trend for each population was fitted by a sim-
ple linear regression under the assumption of normally distrib-
uted errors, with the age-standardized incidence as dependent
variable expressed on a logarithmic scale and the calendar
year as independent variable: lnl
(t) = a
+ b
t, where l
(t) de-
note the age-standardized incidence predicted at year t for
population i; the intercept a
is different for each population,
and b
is the population specific regression coefficient (the
trend), respectively. In such a multiplicative model the regres-
sion coefficient ( ´ 100 %) is a percentage, approximately be-
ing the average relative increase in incidence per year. The
multiplicative regression model was used because it fitted the
data well. It is commonly used in estimating time trends in inci-
dence and allows a simple interpretation of the regression co-
The overall estimate of the relative annual increase was ob-
tained by using a pooled, centralized data set: to start, for each
population the logarithms of the age-standardized incidence
rates and the time points were centralized to make the differ-
ent lengths of the studies and incidence levels more compara-
ble. Then, using the method of least squares, a straight line
constrained to cross the origin of the centralized coordinate
system was fitted to the pooled data set. The regression coeffi-
cient has the same interpretation as in the population-wise
analysis. The analysis was subsequently repeated as weighted
regression, where the residual sum of squares was weighted
with the number of cases in individual studies, to give more
weight for observations with a higher number of cases.
The association between the level of incidence and increase
in incidence was assessed by calculating the correlation coeffi-
cient between the logarithms of incidences in 1983 predicted
by the model and the incidence increases estimated by the mul-
tiplicative model. The year 1983 was chosen because almost all
studies covered it.
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes1396
Predictions until the year 2010 have been made with both
multiplicative and additive regression models, meaning that
the curve produced by fitting the model to the data is simply
extrapolated to the year 2010. In essence, the multiplicative
model fits an exponential curve to the incidence, whereas the
additive model fits a straight line. The additive regression
model is used here to point out the differences between predic-
tions when using alternative models.
Incidence. The mean incidence of Type I diabetes
among the study populations varied from 0.5 to 30.3
per 100 000 a year during the observation period (Ta-
ble 1). The mean incidences were divided into quar-
tiles: low incidence, less than 6.4 per 100 000 a year
(the lowest 25% of the mean incidences), intermedi-
ate 6.4±14.6 per 100 000 a year (50% of the mean in-
cidences, thus the two intermediate groups com-
bined), and high, more than 14.6 per 100 000 a year
(highest 25% of the mean incidences).
Increase in the incidence of Type I diabetes. The rela-
tive change (% per year) in incidence among individu-
al populations ranged from ±0.2% in Colorado (USA)
to 9.5% in Leicestershire (UK) (Table 2). A statisti-
cally significant increase in incidence was found in
65% (24/37) of the populations. An upward tendency
in incidence not reaching statistical significance was
observed in another 12 populations. Only in one pop-
ulation, Colorado (USA), the trend was slightly, but
not significantly, negative (±0.2% per year) and the
upper limit of the 95% confidence interval show that
an increase of 2.2% per year was possible.
The global trend and the annual increase in the in-
cidence of Type I diabetes were estimated from the
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes 1397
Table 1. The studies included in the analysis
Country: area Study
of study
of cases
Degree of
Algeria: Oran 1979±1988 10 0±14 173
4.7 P, R ? F 36
Australia: West 1985±1992 8 0±14 350 14.9 P 99.6% T 37
Austria 1979±1993 15 0±14 1551 7.8 P, R > 90% T 15, 38
Bulgaria, East 1974±1995 22 0±14 818 6.3 R, P 98.8% T 39
Canada: Montreal 1971±1985 15 0±14 839 9.3 R 94% T 40
Canada: Prince 1975±1986 19 0±14 115 23.5 P 99% T 41, 42
Edward Island 1990±1993
China: Shanghai 1980±1993 14 0±14 67 0.7 R 85.2% F 43, 44
East-Germany 1960±1989 30 0±19 9581 6.7 P 99% F 45
Estonia 1983±1996 14 0±14 523 10.2 P 92%
T 42, 46
Finland 1965±1996 32 0±14 9047 30.3 P, R 100% T 14, 47
France 1988±1995 8 0±19 1439 8.0 P 96% T 48
Hungary 1978±1987 10 0±14 1060 6.1 P 96.2% T 49
Iceland 1970±1988 19 0±14 120 9.0 P, R 100% F 50
Israel: Yemenite Jews 1965±1993 29 0±17 1665 5.0 P ? F 51
Italy: Turin 1984±1991 8 0±14 227 8.4 R, P 99% F 52
Japan: Hokkaido 1973±1992 20 0±14 396 1.7 P 100% F 53
Latvia 1983±1996 10 0±14 652 7.2 P 80±100% F, T 54, 55
Libya 1981±1990 10 0±14 251 8.7 P ? T 56
Lithuania 1983±1992 10 0±14 215 6.4 P 95±100% F 54
Malta 1980±1996 17 0±14 90 14.7 P ? F 57
New Zealand: Auckland 1977±1987 11 0±15 206 10.1 P 100% T 58
New Zealand: Canterbury 1982±1990 9 0±19 123 12.7 P 100% F 59
Norway 1973±1982 10 0±14 1914 20.8 R 99.4% F 60
Peru: Lima 1985±1994 10 0±14 111 0.5 P 85 % T 61
Poland: Krakow 1987±1994 8 0±14 396 5.9 P 100% T 62
Poland: Rzeszów 1980±1992 13 0±14 122 5.1 P, R 99% F 63
Poland: Wielkopolska 1970±1985 15 0±16 451 4.4 P > 95% T 15, 64
Slovakia 1985±1995 11 0±14 1127 7.5 P 95% T, F 65, 66
Sweden 1978±1992 15 0±14 5831 24.9 P, R 100 % F 3
UK: Leicestershire 1965±1981 17 0±14 248 7.8 P, R 97.6% F 15
UK: Oxford 1985±1995 11 0±14 1037 18.5 P ? T 67
UK: Plymouth 1975±1996 22 0±14 488 14.9 P 95.3% F 68
UK: Scotland 1976±1993 18 0±14 4182 21.6 R > 95% T 15
UK: Yorkshire 1978±1992 15 0±14 1721 14.3 P 97.6% F 69, 70
USA: Allegheny County 1965±1985 21 0±14 1041 14.7 P, R 100 % T 71
USA: Colorado 1978±1988 11 0±17 1376 7.8 P, R 93.3% F 72
USA: Hawaii 1980±1990 11 0±14 113 12.3 P 97% F 73
P = prospective, R = retrospective
F = figure, T = table
non Estonian
age-standardized incidence rates using the log-linear
regression (Table 2). The p value of less than 0.05 for
the two-sided test for a non-zero regression coeffi-
cient was regarded as evidence for the trend. The
global annual increase was 3.0% (95% CI 2.59; 3.33,
p = 0.0001) during 1960 to 1996, showing a highly sig-
nificant increasing trend. When the annual incidence
rates were weighted with the number of cases in
each individual study, the increase in incidence was
2.5% (95% CI 2.32; 2.66; p = 0.0001). The estimated
population-wise regression lines illustrate well the in-
creasing trends (Fig.1, 2).
Comparison of increase rates. The relation between
the increase in incidence of Type I diabetes and the
average level of incidence expressed as the incidence
in 1983 estimated from the regression model is shown
in Figure 3. There was a significant inverse associa-
tion between the increase and the logarithm of the
level of incidence (r = ±0.56, p = 0.0004). The associa-
tion indicates that the relative increase was more pro-
nounced in the populations with a low incidence.
Nevertheless, in the five populations with the highest
incidence the increase was also statistically signifi-
cant, varying from 1.2 to 3.2% per year.
The incidence level of Type I diabetes and its in-
crease seemed to be similar in some geographically
adjacent populations. For example in the northern
European countries; Finland, Sweden and Norway
where the incidence of Type I diabetes has been high
for a long time, the increase was 1 to 3% per year.
Adjacent countries around the Baltic Sea, Estonia,
Latvia, Lithuania and Poland with an intermediate
or low incidence (4±10 per 100 000 a year) showed
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes1398
Table 2. Relative increase in incidence of Type I diabetes in
children aged 14 years or less in 37 populations. The popula-
tions are arranged in descending order according to the rela-
tive increase per year. The 95% confidence interval is given in
the parentheses. P value stands for the two-sided test for a non-
zero regression coefficient
Country: area Mean incidence Increase in incidence %
per year (95% CI)
p value
UK: Leicestershire 7.8 9.5 (6.51; 12.53) 0.0001
Hungary 6.1 8.5 (6.50; 10.42) 0.0001
Algeria: Oran 4.7 7.9 (1.85; 14.00) 0.0338
USA: Hawaii 7.8 7.8 (1.80; 14.87) 0.0315
Peru: Lima 0.5 7.7 (0.97; 16.40) 0.1197
China: Shanghai 0.7 7.4 (2.30; 12.51) 0.0148
Poland: Krakow 5.9 6.8 (2.27; 11.41) 0.0262
New Zealand: Auckland 10.1 6.4 (4.20; 8.52) 0.0003
Australia: West 14.9 6.3 (2.11; 10.50) 0.0259
Libya 8.7 6.3 (0.69; 11.8) 0.0589
Japan: Hokkaido 1.7 5.9 (4.14; 7.63) 0.0001
Slovakia 7.5 5.5 (3.64; 7.41) 0.0003
Poland: Wielkopolska 4.4 4.8 (1.94; 7.66) 0.0054
France 8.0 3.9 (2.85; 4.94) 0.0003
UK: Oxford 18.5 3.7 (1.82; 5.50) 0.0037
Canada: Prince Edward Island 23.5 3.2 (0.33; 6.38) 0.0728
Israel: Yemenite Jews 5.0 3.2 (2.51; 3.88) 0.0001
Norway 20.8 3.2 (1.19; 5.22) 0.0143
Austria 7.8 2.7 (1.58; 3.76) 0.0003
UK: Plymouth 14.6 2.7 (0.91; 4.50) 0.0079
New Zealand: Canterbury 12.7 2.7 (0.05; 10.50) 0.5262
UK: Scotland 21.6 2.5 (1.85; 3.08) 0.0001
East-Germany 6.7 2.4 (1.96; 2.90) 0.0001
Finland 30.3 2.3 (1.98; 2.57) 0.0001
Iceland 9.0 2.3 (2.38; 6.96) 0.3498
Italy: Turin 8.4 2.2 (3.99; 8.35) 0.5150
Bulgaria: East 6.3 2.1 (1.03; 3.15) 0.0010
UK: Yorkshire 14.3 1.9 (0.30; 3.53) 0.0372
Canada: Montreal 9.3 1.6 (0.67; 3.82) 0.1937
USA: Allegheny County 14.7 1.5 (0.21; 2.83) 0.0348
Sweden 24.9 1.2 (0.42; 2.02) 0.0103
Poland: Rzeszów 5.1 1.1 (3.25; 5.40) 0.6339
Lithuania 6.4 1.1 (4.25; 6.41) 0.7023
Latvia 7.2 0.9 (1.90; 3.75) 0.5350
Malta 14.7 0.5 (2.15; 3.19) 0.7078
Estonia 10.2 0.4 (0.96; 1.76) 0.5741
USA: Colorado 12.3 0.2 (2.52; 2.19) 0.8938
Change globally 3.0 (2.59; 3.33) 0.0001
Change globally (weighted by number of cases) 2.5 (2.32; 2.66) 0.0001
an upward course but not a statistically significant
trend in incidence. The increase in incidence in east-
ern Europe varied from 2.1% per year in East Bul-
garia to 8.5% per year in Hungary. In the United
Kingdom the mean incidence ranged from 14.3 to
21.6 and the increase in incidence ranged between
1.9 and 3.7% except for Leicestershire where the
mean incidence was 7.8 with an increase of 9.5%.
The data for Leicestershire, however, were consider-
ably older (from 1965 to 1981) than from other UK
study populations.
Predictions until the year 2010. Since no effective pre-
vention has thus far been invented or is foreseen in
the near future, we used the observed trends to pre-
dict the incidence of Type I diabetes at least until the
year 2010 (Table 3). For the prediction we applied
both linear and exponential models, since the model
of the increase in incidence is not known. The expo-
nential predictions were only calculated for popula-
tions with a study period of at least 14 years. In gener-
al, the linear model produces more conservative pre-
dictions than the exponential model, however, in
those populations where an increase had started dur-
ing the very last years of the observation period, the
exponential model gave more conservative predic-
tions than the linear.
The predictions based on the linear model show
that Finland will still have the highest incidence in
the world (50 per 100 000 a year) in the year 2010, fol-
lowed by Norway, Prince Edward Island (Canada),
western Australia, Scotland (UK), Oxford (UK),
and Sweden. Despite the large relative increases in
the incidence observed in China and Peru, the abso-
lute incidence rates in these countries would still re-
main low, less than 2 per 100 000 a year. Based on
these predictions, in Japan the incidence will be lower
than 5 per 100 000 a year and in Poland, Latvia and
Lithuania the incidence will be under 10 per 100 000
a year.
There are also other populations in which an increase
in the incidence of Type I diabetes has been recently
reported such as Croatia, Denmark, Kuwait, the
Netherlands, Russia and Switzerland [17±23]. These
studies, however, did not meet the inclusion criteria
of this study and were not included in our analysis.
To find out whether the rising incidence is really a
global phenomenon, we carried out an analysis of in-
cidence trends among 37 populations worldwide for
which the data had been collected for 8 years or
more. The incidence of Type I diabetes is globally in-
creasing by 3.0% per year (or by 2.5%, when the in-
cidences were weighted by the number of diabetic
children included in the individual studies). Confi-
dence intervals for these estimates were fairly narrow
indicating that these estimates are reliable. Accord-
ing to this estimate, the incidence of Type I diabetes
will be 40% higher in 2010 than in 1998. This is a real-
istic, although a rather frightening, scenario.
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes 1399
Fig. 1. Trends in incidence of Type I diabetes in European pop-
ulations. AUT: Austria; BGR: Bulgaria; DEU: East-Germany;
EST: Estonia; FIN: Finland; FRA: France; HUN: Hungary;
ICE: Iceland; ITA: Italy (Turin); LVA: Latvia; LTU: Lithuania;
MAT: Malta; NOR: Norway; POLk: Poland (Krakow); POLr:
Poland (Rzeszów); POLw: Poland (Wielkopolska); SVK: Slo-
vakia; SWE: Sweden; UNKl (UK, Leicestershire); UNKo (Ox-
ford); UNKp (Plymouth); UNKs (Scotland); UNKy (York-
shire). The model fitted to the incidence data was a multiplica-
tive regression model with logarithm of the age standardized
incidence as dependent variable, thus the scale of the incidence
is logarithmic when straight lines were used in drawing the re-
gression lines
Fig. 2. Trends in the incidence of Type I diabetes mellitus in
Non-European populations. DZD: Algeria (Oran); AUSw:
Australia (West); CANm: Canada (Montreal); CANp: Canada
(Prince Edward Island); CHN: China (Shanghai); ISR: Israel
(Yemenite Jews); JPN: Japan (Hokkaido); LIY: Libya; NEZa:
New Zealand (Auckland); NEZc: New Zealand (Canterbury);
PER: Peru (Lima); USAa: USA (Allegheny County); USAc:
USA (Colorado); USAh: USA (Hawaii)
The global variation in the incidence of Type I dia-
betes is prominent [10, 13, 14]. It reflects the distribu-
tion of ethnic diversity showing the importance of the
differential genetic susceptibility among populations.
The incidence is higher among Caucasoid popula-
tions than among Mongoloids and Blacks. Within
ethnic groups, however, there are geographical differ-
ences in incidence depending on the admixture be-
tween racial groups and possible environmental ex-
posures [13]. Although most of the populations in-
cluded in this analysis were Caucasoid, statistically
significant increases in incidence were also found
among the Asian populations in China and Japan,
Mestizos in Peru and also among the Polynesians in
Hawaii. In this literature review it was not possible
to account for ethnic differences within populations
because the authors of the original papers had usually
not given detailed information on incidence in differ-
ent ethnic groups. Where the ethnic groups had been
analyzed separately, however, an increase in the inci-
dence had been observed in all groups, but the rate
of increase could vary from one ethnic group to an-
other. Overall, the increase in the incidence of Type
I diabetes does not seem to be restricted to any par-
ticular ethnic group.
In most countries with a low incidence the stan-
dardized incidence data have been collected during a
relatively short period, which may in some cases ex-
plain the large relative change in incidence. The inci-
dence of Type I diabetes possibly has been underesti-
mated in earlier studies because of incomplete case-
ascertainment and death from undiagnosed diabetes.
Among those populations where the study period
was 18 years or more the increase in incidence was
usually low (from 1.5 to 3.2% per year except Japan,
Hokkaido, 5.9%). Therefore, results from several in-
dividual populations showing large increases should
be interpreted cautiously when the number of cases
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes1400
Fig. 3. Association between the increase in incidence and the
level of incidence of Type I diabetes. The correlation coeffi-
cient between the log-transformed level of incidence predicted
for 1983 and the increase in incidence was ±0.56 (p = 0.0004)
for all populations. The increase in incidence was estimated
from multiplicative regression models. For each population
the level of incidence was calculated as the model-predicted
(fitted) incidence in 1983 according to the multiplicative re-
gression model. Leicestershire, UK, has been excluded. DZD:
Algeria (Oran); AUSw: Australia (West); AUT: Austria;
BGR: Bulgaria; CANm: Canada (Montreal); CANp: Canada
(Prince Edward Island); CHN: China (Shanghai); DEU: East-
Germany; EST: Estonia; FIN: Finland; FRA: France; HUN:
Hungary; ICE: Iceland; ISR: Israel (Yemenite Jews); ITA: Ita-
ly (Turin); JPN: Japan (Hokkaido); LVA: Latvia; LIY: Libya;
LTU: Lithuania; MAT: Malta; NEZa: New Zealand (Auck-
land); NEZc: New Zealand (Canterbury); NOR: Norway;
PER: Peru (Lima); POLk: Poland (Krakow); POLr: Poland
(Rzeszów); POLw: Poland (Wielkopolska); SVK: Slovakia;
SWE: Sweden; UNKo: UK (Oxford); UNKp: UK (Plymouth);
UNKs: UK (Scotland); UNKy: UK (Yorkshire); USAa: USA
(Allegheny County); USAh: USA (Hawaii); USAc: USA
Table 3. Predicted incidence (100000 per year) of Type I dia-
betes in children aged 14 years or less by the year 2010 accord-
ing to both the multiplicative and additive regression model.
The populations are arranged in descending order according
to the predicted incidence in 2010 based on the additive model.
Only the populations with a study period covering 14 years or
more were included in predictions with the multiplicative mod-
Country: area Predicted incidence per
100000/year in 2010
Finland 50.2 57.9
Norway 41.8
Canada: Prince Edward Island 39.2 51.4
Australia: West 36.7
UK: Scotland 34.9 40.0
UK: Oxford 33.0
Sweden 32.2 33.7
New Zealand: Auckland 27.7
New Zealand: Canterbury 25.0
UK: Plymouth 23.4 27.0
USA: Allegheny County 22.5 24.5
USA: Hawaii 21.4
UK: Yorkshire 21.0 23.0
Hungary 20.1
Libya 19.0
Malta 16.8 15.8
Iceland 16.1 16.0
Slovakia 16.0
Algeria: Oran 15.2
Canada: Montreal 14.8 15.0
France 13.7
Poland: Krakow 13.0
Italy: Turin 12.8
Austria 12.8 14.6
East Germany 12.0 15.4
USA: Colorado 12.0
Poland: Wielkopolska 11.8 20.2
Estonia 10.9 11.0
Bulgaria: East 10.3 10.7
Israel: Yemenite Jews 10.1 12.9
Latvia 8.9
Lithuania 7.5
Poland: Rzeszów 6.7
Japan: Hokkaido 4.1 7.9
China: Shanghai 1.7 3.3
Peru: Lima 1.3
is small and the study period short. The analysis of
the pooled data was repeated excluding populations
for which the case ascertainment level was not report-
ed (Algeria, Israel, Libya, Malta and Oxford, UK).
The results were essentially the same as in the analy-
sis using all data: 2.95% per year vs 2.96% per year,
respectively, and for the weighted regression, 2.40%
vs 2.49%, respectively.
There is presently no way to know whether the ob-
served trend in incidence might reflect a change in
the age at onset of diabetes instead of a true rise in
prevalence. The increase in incidence in 0±14-year-
olds might just be a transition of the age at onset
from the age group 15 years or older. Data on inci-
dence trends in older age groups exist from just a
few populations; thus, reliable information a possible
decrease in incidence in young adults is not available.
Our main result is that the incidence is globally in-
creasing in the age group of 0±14-year-olds.
The genetic factors have been shown to be impor-
tant in the susceptibility to Type I diabetes [5, 6]. Al-
though it is possible that the part of the population
genetically predisposed for Type I diabetes is increas-
ing, this increase may have been modest during the
last decades and not alone a sufficient cause for the
observed increase in incidence. The changes in the
genetic code of the human populations are usually
slow. In this analysis even the longest study period
covered only 30 years, which is approximately the
time span of one generation. It is very unlikely that a
three to tenfold increase in the proportion of subjects
with genetic susceptibility to Type I diabetes has tak-
en place in any population during such a short time.
Instead, the penetrance of the susceptibility genes
might be changing. The penetrance is likely to be de-
termined by an interaction between several suscepti-
bility genes and unknown environmental factors [4,
During the recent years much attention has been
paid to the identification and possible control of envi-
ronmental factors which possibly initiate or trigger
the process leading to Type I diabetes. Although
some studies suggest associations between environ-
mental factors such as diet and viral infections with
the risk of Type I diabetes [25±35], their causative
role has not been shown. It is also difficult to show
that any of these environmental factors has changed
in such a way that a continuous global increase in the
incidenceof TypeI diabeteswould beeasily explained.
The incidence of Type I diabetes is increasing
worldwide. Thus far no population has been identi-
fied in which the incidence has significantly de-
creased. The population-based WHO DIAMOND
Project and the EURODIAB study started at the be-
ginning of the 1990s but have not yet reported results
from the long-term progress in the incidence of Type
I diabetes. It seems obvious that in both of these stud-
ies the 10-year monitoring period planned thus far is
too short to produce reliable trend estimations and
predictions for the change of the incidence of Type I
diabetes, especially in countries where incidence is
low. There is a need to continue with the communi-
ty-based registries on Type I diabetes worldwide. Ef-
forts are also needed to identify effective primary
prevention measures for Type I diabetes to stop the
global increase in the incidence of this disease.
Acknowledgements. We thank the Diabetes Epidemiology Re-
search International (DERI) Group for the data for several
countries used in this study.
1. Cudworth AG, Wolf E, Gorsuch AN, Festenstein H (1979)
A new look at HLA genetics with particular reference to
type-1 diabetes. Lancet 2: 389±391
2. Tuomilehto-Wolf E, Tuomilehto J (1991) HLA antigens in
insulin-dependent diabetes mellitus. Ann Med 23: 481±
3. Dahlquist G, Mustonen L (1994) Childhood onset diabetes
± time trends and climatological factors. Int J Epidemiol 23:
4. Todd JA (1997) Genetics of type 1 diabetes. Pathol Biol
(Paris) 45: 219±227
5. Kaprio J, Tuomilehto J, Koskenvuo M et al. (1992) Concor-
dance for type 1 (insulin-dependent) and type 2 (non-insu-
lin-dependent) diabetes mellitus in a population-based co-
hort of twins in Finland. Diabetologia 35: 1060±1067
6. Kyvik KO, Green A, Beck-Nielsen H (1995) Concordance
rates of insulin dependent diabetes mellitus: a population
based study of young Danish twins. BMJ 311: 913±917
7. Blom L, Nyström L, Dahlquist GG (1991) The Swedish
childhood diabetes study: vaccinations and infections as
risk determinants for diabetes in childhood. Diabetologia
34: 176±181
8. Virtanen SM, Räsänen L, Aro A et al. (1992) Feeding in in-
fancy and the risk of type 1 diabetes mellitus in Finnish
children. Diabet Med 9: 815±819
9. LaPorte RE, Tajima N, kerblom HK et al. (1985) Geo-
graphic differences in the risk of insulin-dependent diabe-
tes mellitus: the importance of registries. Diabetes Care
8[Suppl 1]:S101±S107
10. Rewers M, LaPorte RE, King H, Tuomilehto J (1988)
Trends in the prevalence and incidence of diabetes: insu-
lin-dependent diabetes mellitus in childhood. World
Health Stat Q 41: 179±189
11. WHO DIAMOND Project Group on Epidemics (1992)
Childhood diabetes, epidemics, and epidemiology: an ap-
proach for controlling diabetes. Am J Epidemiol 135:
12. Green A, Gale EA, Patterson CC for the EURODIAB
ACE study (1992) Incidence of childhood-onset insulin-de-
pendent diabetes mellitus: the EURODIAB ACE study.
Lancet 339: 905±909
13. Karvonen M, Tuomilehto J, Libman I, LaPorte R for the
World Health Organization DIAMOND Project Group
(1993) A review of the recent epidemiological data on the
worldwide incidence of type 1 (insulin-dependent) diabetes
mellitus. Diabetologia 36: 883±892
14. Karvonen M, Pitkäniemi M, Pitkäniemi J, Kohtamäki K,
Tajima N, Tuomilehto J for the World Health Organization
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes 1401
DIAMOND Project Group (1997) Sex difference in the in-
cidence of insulin-dependent diabetes mellitus: an analysis
of the recent epidemiological data. Diabetes Metab Rev
13: 275±291
15. Diabetes Epidemiology Research International Group
(1990) Secular trends in incidence of childhood IDDM in
10 countries. Diabetes 39: 858±864
16. Diabetes Epidemiology Research International Group
(1988) Geographic patterns of childhood insulin-depen-
dent diabetes mellitus. Diabetes 37: 1113±1119
17. Schoenle EJ, Molinari L, Bagot M, Semadeni S, Wiesen-
danger M (1994) Epidemiology of IDDM in Switzerland.
Increasing incidence rate and rural-urban differences in
Swiss men born 1948±1972. Diabetes Care 17: 955±960
18. Ruwaard D, Gijsen R, Bartelds IM, Hirasing RA, Verkleij
H, Kromhaut D (1996) Is the incidence of diabetes increas-
ing in all age-groups in The Netherlands? Results of the
second study in the Dutch Sentinel Practice Network. Dia-
betes Care 19: 214±218
19. Bingley PJ, Gale EA (1989) Rising incidence of IDDM in
Europe. Diabetes Care 12: 289±295
20. Green A, Andersen PK, Svendsen AJ, Mortensen K (1992)
Increasing incidence of early onset type 1 (insulin-depen-
dent) diabetes mellitus: a study of Danish male birth co-
horts. Diabetologia 35: 178±182
21. Jaksic J, Matic I, Stojnic E, Juros A, Pelajic A (1996) Inci-
dence of insulin dependent diabetes mellitus in children
aged 0±19 in the Sibenic area. Diabetol Croat 29±33
22. Choubnikova J, Shubnikof E, Kalashnikova L (1996) The
increase of type 1 diabetes incidence among children in
Novosibirsk city. 32nd Annual Meeting of the European
Association for the Study of Diabetes, Vienna (Abstract)
23. Shaltout AA, Qeabazard MA, Abdella et al. (1995) High
incidence of childhood-onset IDDM in Kuwait. Diabetes
Care 18: 923±927
24. Cordell HJ, Todd JA (1995) Multifactorial inheritance in
type 1 diabetes. Trends Genet 11: 499±504
25. Borch-Johnsen K, Joner G, Mandrup-Poulsen T et al.
(1984) Relation between breast-feeding and incidence
rates of insulin-dependent diabetes mellitus. A hypothesis.
Lancet 2: 1083±1086
26. Kostraba JN, Cruickshanks KJ, Lawler-Heavner J et al.
(1993) Early exposure to cow's milk and solid foods in in-
fancy, genetic predisposition, and risk of IDDM. Diabetes
42: 288±295
27. Elliott RB, Martin JM (1984) Dietary protein: a trigger of
insulin-dependent diabetes in the BB rat? Diabetologia
26: 297±299
28. Helgason T, Jonasson MR (1981) Evidence for a food addi-
tive as a cause of ketosis-prone diabetes. Lancet 2: 716±720
29. Dahlquist GG, Blom LG, Persson LA, Sandström AI, Wall
SG (1990) Dietary factors and the risk of developing insulin
dependent diabetes in childhood. BMJ 300: 1302±1306
30. Virtanen SM, Jaakkola L, Räsänen L et al. (1994) Nitrate
and nitrite intake and the risk for type 1 diabetes in Finnish
children. Diabet Med 11: 656±662
31. Rayfield EJ, Seto Y (1978) Viruses and the pathogenesis of
diabetes mellitus. Diabetes 27: 1126±1140
32. Gamble DR (1976) A possible virus etiology for juvenile
diabetes. In: Creutzfeld W, Köbberlin J, Neel V (eds) The
genetics of diabetes mellitus. Springer, Berlin Heidelberg
New York, pp 95±105
33. Hyöty H, Leinikki P, Reunanen A et al. (1988) Mumps in-
fections in the etiology of type 1 (insulin-dependent) diabe-
tes. Diabetes Res 9: 111±116
34. Hyöty H, Hiltunen M, Knip M et al. (1995) A prospective
study of the role of coxsackie B and other enterovirus in-
fections in the pathogenesis of IDDM. Diabetes 44:
35. Hiltunen M, Hyöty H, Karjalainen J et al. (1995) Serologi-
cal evaluation of the role of cytomegalovirus in the patho-
genesis of IDDM: a prospective study. Diabetologia 38:
36. Bessaoud K, Boudraa G, Deschamps I, Hors J, Benbouab-
dallah M, Touhami M (1990) Epidemiologie du diabte in-
sulinodØpendant juvØline en AlgØrie (Wilaya D'Oran).
Rev Epidemiol Sante Publique 38: 91±99
37. Kelly HA, Russell MT, Jones TW, Byrne GC (1994) Dra-
matic increase in incidence of insulin dependent diabetes
mellitus in Western Australia. Med J Aust 161: 426±429
38. Schober E, Schneider E, Waldhör T, Tuomilehto J and
Austrian Diabetes Incidence Study Group (1995) Increas-
ing incidence of IDDM in Austrian children. A nationwide
study 1979±1993. Diabetes Care 18: 1280±1283
39. Tzaneva V, Iotova V, Bruining GJ (1998) Increase in
IDDM incidence in Bulgarian children (1974±1995). J Pe-
diatr Endocrinol Metab 11: 725±732
40. Siemiatycki J, Colle E, Campbell S, Dewar R, Aubert D,
Belmonte MM (1988) Incidence of IDDM in Montreal by
ethnic group and by social class and comparisons with eth-
nic groups living elsewhere. Diabetes 37: 1096±1102
41. Tan MH, Wornell MC, Beck AW (1981) Epidemiology of
diabetes mellitus in Prince Edward Island. Diabetes Care
4: 519±524
42. Väänänen S, Kohtamäki K, Karvonen M, Tuomilehto J for
the Diamond Research Group (1997) Incidence of IDDM
Worldwide Preliminary Analysis of the WHO DIAMOND
Incidence Data. International Diabetes Epidemiology
Group Symposium. Savonlinna, Finland (Abstract)
43. Fu H, Shen SX, Chen ZW et al. (1994) Shanghai, China,
has the lowest confirmed incidence of childhood diabetes
in the world. Diabetes Care 17: 1206±1208
44. Shen SX, Wang HB, Chen ZW et al. (1996) The incidence
of insulin-dependent diabetes mellitus in urban districts of
Shanghai (1989±1993). J Pediatr Endocrinol Metab 9:
45. Michaelis D, Jutzi E, Vogt L (1993) Epidemiology of insu-
lin-treated diabetes mellitus in the East-German popula-
tion: differences in long-term trends between incidence
and prevalence rates. Diabetes Metab 19: 110±115
46. Podar T, LaPorte RE (1993) Incidence of childhood diabe-
tes did not increase in Estonia during 1980±89. Diabetes
Metab 19: 361±363
47. Tuomilehto J, Virtala E, Karvonen M et al. (1995) Increase
in incidence of insulin-dependent diabetes mellitus among
children in Finland. Int J Epidemiol 24: 984±992
48. LØvy-Marchal C, and Groupe du Registre d'incidence
du Diabte Insulino-DØpendant de l'Enfant (1998) Évolu-
tion de l'incidence du diabte insulino-dØpendant de l'en-
fant en France. Rev Epidemiol Sante Publique 46: 157±163
49. SoltØsz G, Madµcsy L, BØkefi D, Dankó I, the Hungarian
Childhood Diabetes Epidemiology Group (1990) Rising
incidence of Type 1 diabetes in Hungarian children
(1978±1987). Diabet Med 7: 111±114
50. Helgason T, Danielsen R, Thorsson AV (1992) Incidence
and prevalence of type 1 (insulin-dependent) diabetes mel-
litus in Icelandic children 1970±1989. Diabetologia 35:
51. Shamis I, Gordon O, Albag Y, Goldsand G, Laron Z (1997)
Ethnic differences in the incidence of childhood IDDM in
Israel (1965±1993) Marked increase since 1985, especially
in Yemenite Jews. Diabetes Care 20: 504±508
52. Bruno G, Merletti F, De Salvia A, Lezo A, Arcari R, Pa-
gano G, Piedmont Study Group for Diabetes Epidemiolo-
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes1402
gy (1997) Comparison of incidence of insulin-dependent di-
abetes mellitus in children and young adults in the province
of Turin, Italy, 1984±91. Diabet Med 14: 964±969
53. Matsuura N, Fukuda K, Okuno A et al. (1998) Descriptive
epidemiology of IDDM in Hokkaido, Japan: the Child-
hood IDDM Hokkaido Registry. Diabetes Care 21:
54. Padaiga Z, Tuomilehto J, Karvonen M et al. (1997) Inci-
dence trends in childhood onset IDDM in four countries
around the Baltic sea during 1983±1992. Diabetologia 40:
55. Brigis G, Robeza I (1997) Childhood onset IDDM inci-
dence rates in Latvia. Outbreak in 1995, comparisons by
ethnicity International Diabetes Epidemiology Group
Symposium. Savonlinna, Finland (Abstract)
56. Kadiki OA, Moawad SE (1993) Incidence and prevalence
of type 1 diabetes in children and adolescents in Benghazi,
Libya. Diabet Med 10: 866±869
57. Schranz AG, Prikatsky V (1989) Type 1 diabetes in the
Maltese Islands. Diabet Med 6: 228±231
58. Elliott RB, Pilcher CC (1985) Childhood diabetes in the
Auckland area. N Z Med J 98: 922±923
59. Scott RS, Brown LJ, Darlow BA, Frobes LV, Moore MP
(1992) Temporal variation in incidence of IDDM in Can-
terbury, New Zealand. Diabetes Care 15: 895±899
60. Joner G, Sùvik O (1989) Increasing incidence of diabetes
mellitus in Norwegian children 0±14 years of age
1973±1982. Diabetologia 32: 79±83
61. SeclØn S, Rojas MI, Nuæez H, Valdivia H, Millones B, Dia-
betes Epidemiology Research Peruvian Group (DERPG)
(1997) Type 1 (insulin-dependent) diabetes in Mestizo chil-
dren of Lima, Peru. Report on a ten years (1985±94) inci-
dence. International Diabetes Epidemiology Group Sym-
posium. Savonlinna, Finland. (Abstract)
62. Szybìnski Z, Czyzyk A, Wasik R, Dziatkowiak H, Cie-
chanowska M, Symonides-Lawecka A, Szurkowska DP
(1997) Epidemiology of IDDM in Poland ± a 10 year per-
spective International Diabetes Epidemiology Group Sym-
posium. Savonlinna, Finland (Abstract)
63. Grzywa MA, Sobel AK (1995) Incidence of IDDM in the
province of Rzeszów, Poland, 0- to 29-year-old age-group,
1980±1992. Diabetes Care 18: 542±544
64. Rewers M, Stone RA, LaPorte RE et al. (1989) Poisson re-
gression modeling of temporal variation in incidence of
childhood insulin-dependent diabetes mellitus in Alleghe-
ny County, Pennsylvania, Wielkopolska, Poland, 1970±
1985. Am J Epidemiol 129: 569±581
65. Michalkovµ DM, Cernay J, Dankovà A, Rusnµk M,
Fandµkova K, Slovak Childhood Diabetes Epidemiology
Study Group (1995) Incidence and prevalence of childhood
diabetes in Slovakia (1985±1992). Diabetes Care 18:
66. Michalkova D, Barak L, Dankova A, Pastor K, Mikulecky
M, Cernay J, Slovak Pediatric Epidemiology Group
(1997) The incidence and prevalence of type 1 diabetes
mellitus in Slovak children from 1985 to 1995. International
Diabetes Epidemiology Group Symposium. Savonlinna,
Finland (Abstract)
67. Gardner SG, Bingley PJ, Sawtell PA, Weeks S, Gale EA,
the Bart's-Oxford Study Group (1997) Rising incidence of
insulin dependent diabetes in children aged under 5 years
in the Oxford region: time trend analysis. BMJ 315:
68. Zhao HX, Stenhouse E, Soper C et al. (1999) Incidence of
childhood onset Type I diabetes mellitus in Devon and
Cornwall, England, 1975±1996. Diabetic Med 16: (in press)
69. Staines A, Bodansky HJ, Lilley HE, Stephenson C, McNal-
ly RJ, Cartwright RA (1993) The epidemiology of diabetes
mellitus in the United Kingdom: The Yorkshire Regional
Childhood Diabetes Register. Diabetologia 36: 1282±1287
70. McKinney PA, Law G, Staines AS, Bodansky HJ (1995) In-
cidence of diabetes in children. BMJ 310: 1672
71. LaPorte RE, Fishbein HA, Drash AL et al. (1981) The
Pittsburgh insulin-dependent diabetes mellitus (IDDM)
registry: the incidence of insulin-dependent diabetes melli-
tus in Allegheny County, Pennsylvania (1965±1976). Dia-
betes 30: 279±284
72. Kostraba JN, Gay EC, Cai Yet al. (1992) Incidence of insu-
lin-dependent diabetes mellitus in Colorado. Epidemiolo-
gy 3: 232±238
73. Patrick SL, Kadohiro JK, Waxman SH et al. (1997) IDDM
Incidence in a Multiracial Population. The Hawaii IDDM
Registry, 1980±1990. Diabetes Care 20: 983±987
P. Onkamo et al.: Worldwide increase in incidence of Type I diabetes 1403
... The incidence of type 1 diabetes is increasing world-wide in both adults [1] and children [2]. Type 1 diabetes is caused by autoimmune destruction of pancreatic β-cells resulting in greatly diminished capacity to produce and secrete insulin [3]. ...
... The findings that type 1 diabetes is increasing worldwide [1,2] and that the role of genetic factors is decreasing with time [10] suggest that environmental factors are playing an increasingly important role in the aetiology of type 1 diabetes. Type 1 diabetes is known to have a long preclinical phase with islet cell auto-antibodies appearing in the early years of life [11,12]. ...
... These findings are consistent with the hygiene hypothesis [14,20] which suggests that lack of microbial exposure in early life predisposes to type 1 diabetes. It might also help explain the reported increase in incidence of type 1 diabetes [1,2,21] since the prevalence of infectious diseases has decreased and sanitation standards have improved over the same period. This also supports the 'old friends hypothesis', which suggests that the depletion of organisms from the urban environment that accompanied the evolution of mammals is one of the reasons for the increasing incidence of chronic inflammatory disorders since the mid-nineteenth century in developed countries [22]. ...
Full-text available
Background The prevalence of type 1 diabetes is increasing worldwide, suggesting that unknown environmental factors are becoming increasingly important in its pathogenesis. Aim The aim of the study was to investigate the possible role of a number of prenatal and perinatal factors in the aetiology of type 1 diabetes. Methods Mothers of patients diagnosed with type 1 diabetes (cases) and mothers of children born on the same day and of the same sex as type 1 diabetes patients (controls) were interviewed on a number of prenatal and perinatal factors of interest. Results Hand washing prior to eating, frequency of bathing and total stress score were found to be positively associated with the development of type 1 diabetes on univariate analyses. Hand-washing prior to eating and frequency of house cleaning were independently associated with an increased risk of type 1 diabetes, whilst getting dirty was associated with a reduced risk in multivariate analyses. There was no association of type 1 diabetes to removing of outdoor shoes indoors or to the age of first attendance to school or pre-school. There were also no significant associations to parental smoking, parental age, birth order, infant feeding, antibiotic use, mode of delivery or birth weight. Conclusion Our data suggest that factors that affect the skin or gut microbiome might be more important than infections or factors affecting the microbiome at other sites.
... It affects 18-20 per 100,000 children a year in the UK. [Onkamo et al, 1999] Globally, incidences of Type 2 diabetes are increasing, for example reported' ...
p>Continuous Subcutaneous Insulin Infusion (CSII) represents the latest technology in insulin delivery and, whilst demanding, arguable provides additional QoL benefits to other insulin regimens. Existing literature provides mixed reports regarding QoL benefits associated with CSII. However, poor methodology, small samples and inappropriate measures may explain this ambiguity. Due to the subjective nature of QoL, quantitative measures alone may be insufficient to capture the impact on QoL. This thesis supports the conclusion that QoL benefits may be observed in association with CSII therapy. Both qualitative and quantitative studies demonstrate improved QoL (e.g. increased independence, freedom and flexibility, particularly in terms of meal timing and content) for CSII users (measured using a range of generic, health related and diabetes-specific measures). These findings are consistent for children, adolescents and adults using CSII. In addition, Chapters 4 and 6 report QoL benefits for members of the family as well as the individual with diabetes. Specific life domains important for QoL cited by children and adolescents are consistent with those in the literature, i.e. ‘family’, ‘friends’ and ‘school’, while parents most frequently cited ‘health’, ‘family’ and ‘work’ as important for their QoL. All domains were rated as better since commencement of CSII therapy. Future research needs to focus on identifying the factors that predict success using CSII therapy and the individuals who would most be suited to benefit from using CSII therapy.</p
... The threat for fathers has become intricate compared to mothers and could be elevated if the first affected individual in lineage turned into detected earlier age 7 (Harjutsalo et al., 2006). The prevalence of DM1 is rising globally (Onkamo et al., 1999). Only minimal information is debunked that specifically which environmental factors contribute in development of diabetes type 1. ...
Full-text available
Type 1 diabetes (T1D) with a solid genomic factor, is a multifactorial disease. which arise through autoimmune annihilation of pancreatic β cells. Epidemiologic forms of T1D type 1 diabetes by cultural, demographic, geographic, biological and some additional aspects in population are offered to increase awareness with regard to risks, past linkages, etiologic and impediment of DM1. Informatio n after huge epidemiological trainings indicate internationally that the occurrence of type 1 diabetes T1D increased globally by 2-5% and the occurrence of T1D is around 1 in 300, by 18 years of phase in the US. Study on hazard aspects for type 1 diabetes T1D is a dynamic part of study to classify inherited and conservational reasons that might be theoretically targeted for interference. Though important developments been completed in the experimental maintenance of type 1 diabetes T1D with experimental consequences and resulting developments in class of natural life, significant extra requirements to improve care of and eventually discover a medication for type 1 diabetes T1D. Epidemiological trainings take a significant ongoing part to examine the clinical care, prevention, complex causes and therapy of type 1 diabetes T1D. INS, HLA CTLA4 and PTPN22 are measured to be established by type 1 diabetes (T1D) vulnerability genetic factor. CTLA4, PTPN22 and HLA are identified to be complicated in protected instruction. Hypothesis mostly recognized vulnerability genetic factor appear to almost increase with other loci on the risk of disease including the joint effect of PTPN22 and HLA. The combined outcome of many vulnerability loci discussed the actual risk of t ype 1 diabetes but also applies to the same unimportant part of the overall population. By means of numerous vulnerability genotypes associated with HLA genotypic factor appeared to slightly effect the prediction of disease.
... The incidence of T1D in the pediatric population has an increasing trend worldwide. The rate of change in T1D incidence varies depending on the ethnic origin, geographical region and industrialization status (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18). ...
Objective: Type 1 Diabetes mellitus (T1D) incidence in children has an increasing trend with a variable rate depending on region and ethnicity. Our group had firstly reported T1D incidence in Diyarbakır in the year 2011. The present study aims to evaluate the current incidence rate of pediatric T1D in Diyarbakır, and compare the incidence, clinical and presenting characteristics of the cases with those reported in our first report. Methods: Hospital records of the patients under 18 years old and diagnosed with T1D in Diyarbakır city between 1st January 2020 and 31st December 2020 were retrieved, and their medical data was extracted. Demographic population data were obtained from address-based census records of the Turkish Statistical Institution (TSI). Results: Fifty-seven children and adolescents were diagnosed with T1D. Of those, 34 were female (59.6%), indicating a male/female ratio of 1.47. The mean age of diagnosis was 9.5±3.9 (0.8-17.9). According to the data obtained from TSI, the population between the ages of 0-18 was found to be 709803. T1D incidence was calculated as 9.14/105 in the 0-14 age group and 8.03/105 in the 0-18 age group. The cumulative increase in the incidence of T1D in the 0-14 age group was 26.9% suggesting an increasing rate of 2.7% per year. The frequency of presentation with DKA was 64.9%.
... The number of autoimmune disorders (ADs) affecting at least 5% of individuals vaccinated in childhood has increased significantly within the last 30 years worldwide [3][4][5]. The question was even raised as to whether vaccination should or should not be recommended for those with a personal or family history of an AD [6]. ...
Full-text available
Vaccination as an important tool in the fight against infections has been suggested as a possible trigger of autoimmunity over the last decades. To confirm or refute this assumption, a Meta-analysis of Autoimmune Disorders Association With Immunization (MADAWI) was conducted. Included in the meta-analysis were a total of 144 studies published in 1968–2019 that were available in six databases and identified by an extensive literature search conducted on 30 November 2019. The risk of bias classification of the studies was performed using the Newcastle–Ottawa Quality Assessment Scale. The strength of evidence was assessed using the Grading of Recommendations Assessment, Development, and Evaluation. While our primary analysis was conducted in terms of measures of association employed in studies with a low risk of bias, the robustness of the MADAWI outcome was tested using measures independent of each study risk of bias. Additionally, subgroup analyses were performed to determine the stability of the outcome. The pooled association of 0.99 (95% confidence interval, 0.97–1.02), based on a total of 364 published estimates, confirmed an equivalent occurrence of autoimmune disorders in vaccinated and unvaccinated persons. The same level of association reported by studies independently of the risk of bias was supported by a sufficient number of studies, and no serious limitation, inconsistency, indirectness, imprecision, and publication bias. A sensitivity analysis did not reveal any discrepancy in the primary result. Current common vaccination is not the cause of any of the examined autoimmune disorders in the medium and long terms.
... Human EVs are ubiquitous and responsible for serious diseases such as poliomyelitis, myocarditis and aseptic meningitis [99]. However, many EV infections cause subclinical or mild disease and are thus underreported, with a small proportion proceeding to clinical identification [100]. More severe EV infection is typically seen in children and neonates, with proposed intrinsic immunity in the adult mature gut moderating the course of infection and preventing viraemia [101]. ...
Full-text available
For over a century, viruses have left a long trail of evidence implicating them as frequent suspects in the development of type 1 diabetes. Through vigorous interrogation of viral infections in individuals with islet autoimmunity and type 1 diabetes using serological and molecular virus detection methods, as well as mechanistic studies of virus-infected human pancreatic β-cells, the prime suspects have been narrowed down to predominantly human enteroviruses. Here, we provide a comprehensive overview of evidence supporting the hypothesised role of enteroviruses in the development of islet autoimmunity and type 1 diabetes. We also discuss concerns over the historical focus and investigation bias toward enteroviruses and summarise current unbiased efforts aimed at characterising the complete population of viruses (the “virome”) contributing early in life to the development of islet autoimmunity and type 1 diabetes. Finally, we review the range of vaccine and antiviral drug candidates currently being evaluated in clinical trials for the prevention and potential treatment of type 1 diabetes.
Full-text available
Diabetes Mellitus is a group of metabolic diseases characterized by hyperglycemia and metabolic disturbances of various metabolisms of Carbohydrates and majorly caused by the Dysfunction of Beta cells in the pancreas. It is of increasing concern, in which a person will have a hyperglycemia, because of the body does not produce enough insulin or because the cells in the body does not respond to the insulin that is produced, majorly, diabetes have been classified into three types. Type I diabetes mellitus results from inability or failure of our body to produce insulin, which increase Blood glucose levels rapidly and whereas in type II diabetes mellitus the person requires to inject the insulin or wear an insulin pump has emerged as a pandemic health problem in the world right now, and the prevalence is increasing rapidly and the type II diabetes mellitus which accounts for about 20% to 50% cases of new-onset of diabetes in the young people, it is a common endocrine disorder, which is also associated with several electrolytic disorders and this also leads to the disturbances in the thyroid gland. The thyroid gland is one of the most important organs in the human body. It regulates the majority of the body's physiological actions. Thyroid hormone has an impact on renal tubular function and the renin-angiotensin system and is associated with hemodynamic and cardiovascular alterations that interfere with renal blood flow. Conversely, the kidney is not only an organ for the metabolism and elimination of TH but also a target organ of some of the actions of the iodothyronines. The thyroid hormone is a central regulator of body functions, disorder of thyroid functions is considered to cause electrolytic disorders in diabetes mellitus patients. The present cross-sectional study investigated whether thyroid parameter concentrations, including thyroid-stimulating hormone (TSH), free triiodothyronine (FT3), free thyroxine (FT4), T3, T4, thyroid peroxidase antibody (TPOAb), and thyroglobulin antibody (TgAb). it had been correlated with electrolytes levels in Diabetic patients. The aim of the study is to study the serum electrolytes levels and thyroid dysfunction in the DM patients. It is a retrospective-cross sectional study. Sample size-100 (50-DM patients and 50-Non Diabetic age matched controls). Medical records of and laboratory reports of 50 patients and 50 controls were retrieved and parameters were retrospectively viewed for clinical findings. The results had showed a significant difference among the controls and the patients. It may be concluded from this study that dysregulation of glucose homeostasis may leads to renal failure, renal stones, cardiac arrhythmias due to increase in sodium & potassium levels, and it also can cause hypogonadism, in future the further studies are needed in order to study the underlying mechanism of electrolytes in diabetes mellitus.
Full-text available
Type 1 diabetes (T1D) is a chronic disease caused by the destruction of pancreatic β cells, which is driven by autoreactive T lymphocytes. It has been described that a high proportion of T1D patients develop other autoimmune diseases (AIDs), such as autoimmune thyroid disease, celiac disease, or vitiligo, which suggests the existence of common etiological factors among these disorders. In this regard, genetic studies have identified a high number of loci consistently associated with T1D that also represent established genetic risk factors for other AIDs. In addition, studies focused on identifying the shared genetic component in autoimmunity have described several common susceptibility loci with a potential role in T1D. Elucidation of this genetic overlap has been useful in identifying key molecular pathways with a pathogenic role in multiple disorders. In this review, we summarized recent advances in understanding the shared genetic component between T1D and other AIDs and discuss how the identification of common pathogenic mechanisms can help in the development of new therapeutic approaches as well as in improving the use of existing drugs.
Aims Dietary intake provides a potential intervention target to reduce the high risk for coronary artery disease (CAD) in type 1 diabetes. This effort aimed to identify patterns of nutrient intake in young/middle-aged adults with type 1 diabetes and to examine associations between those patterns and development of CAD. Methods Principal component analysis was used to derive nutrient intake patterns among 514 individuals with childhood-onset (<17 years old) type 1 diabetes aged 18+ years and free of CAD (defined as CAD death, myocardial infarction, revascularization, ischemia, or study physician diagnosed angina). Cox models were used to assess the association between nutrient patterns and CAD incidence over 30-years of follow-up. Results Three nutrient principal components (PC) were identified: PC1 (high caffeine and saccharin intake), PC2 (high alcohol and caffeine, lower intake of essential nutrients) and PC3 (higher fiber/micronutrients, low alcohol). In unadjusted Cox models, only PC1 (negatively) and PC2 (positively) were associated with CAD risk. These associations were no longer significant after adjusting for diabetes duration. Conclusions Important dietary components underlying the three patterns identified may have been influenced by diabetes duration or age. Future research can continue to explore patterns of nutrient intake over time and CAD development in type 1 diabetes.
A number of animal models of virus-induced diabetes have now been described; Encephalomyocarditis (EMC) virus (4), foot-and-mouth virus (2), and Coxsackie B4 virus (3) have all been found to be diabetogenic in animals. The diabetes described in these reports has usually been mild and transient, and although of great interest, it may not be relevant to the problem of diabetes in man.
Coxsackievirus B infections have been associated with clinical manifestation of insulin-dependent diabetes mellitus (IDDM) in several studies, but their initiating role in the slowly progressing β-cell damage is not known. This is the first prospective study designed to assess the role of coxsackie B and other enterovirus infections in the induction and acceleration of this process. Three separate series were studied: 1) an intrauterine exposure series comprising 96 pregnant mothers whose children subsequently manifested IDDM and 96 control mothers whose children remained nondiabetic; 2) a cohort of 22 initially unaffected siblings of diabetic children who were followed until they developed clinical IDDM (mean observation time, 29 months) and 110 control siblings who remained nondiabetic; 3) a case-control series comprising 90 children with newly diagnosed IDDM and 90 control subjects. Enterovirus infections were identified on the basis of significant increases in serum IgG, IgM, or IgA class antibodies against a panel of enterovirus antigens (capture radioimmunoassay). Enterovirus antibodies were significantly elevated in pregnant mothers whose children subsequently manifested IDDM, particularly in cases in which IDDM appeared at a very young age, before the age of 3 years (P < 0.005). Serologically verified enterovirus infections were almost two times more frequent in siblings who developed clinical IDDM than in siblings who remained nondiabetic (mean, 1.0 vs. 0.6 infections/follow-up year; P < 0.001). This difference was seen both close to the diagnosis of IDDM and several years before diagnosis. Up to 19% (10 of 52) of the infections in prediabetic siblings were associated with increases in islet cell antibody (ICA) levels, and 83% (10 of 12) of ICAs increase with enterovirus infections. The corresponding figures in control siblings were 3% (5 of 185, P < 0.001) and 38% (5 of 13, Ns). IgM class enterovirus antibodies were slightly elevated in young children (<3 years old)with newly diagnosed IDDM (P < 0.05), but not in older patients. These observations suggest that exposures to enterovirus infections, both in utero and in childhood, are able to induce β-cell damage and lead to clinical IDDM after a varying subclinical period.
The intakes of nitrate and nitrite of children and their parents from food and drinking water were estimated in a Finnish nation-wide case-control study on the epidemiology of Type 1 diabetes. The study population consisted of 684 case and 595 control children; 548 case-control pairs of fathers; and 620 case-control pairs of mothers. The consumption frequencies of foods which are important sources of nitrate and nitrite were assessed by structured questionnaire. Nitrate and nitrite concentration data were collected from Finnish water works. Diabetic children's and their mothers' daily dietary intake of nitrite was greater compared with that of control children and mothers (for case and control children 0.9 mg vs 0.8 mg, for case and control mothers 0.9 mg vs 0.8 mg, p<0.001). Case mothers compared with control mothers received less (p<0.05) nitrate from their diet. No differences were observed in the intake of nitrate or nitrite from drinking water. Dietary nitrite intake of children (odds ratios and 95% confidence intervals for the second, third, and fourth quartile 1.16, 0.82–1.65; 1.49, 1.06–2.10; 2.32, 1.67–3.24, respectively) and mothers (odds ratios and 95% confidence intervals for the second, third, and fourth quartile 1.15, 0.76–1.74; 1.29, 0.87–1.91; 1.98, 1.35–2.90, respectively) was positively associated with the risk for Type 1 diabetes independently from length of mother's education, child's or mother's age, place of residence or mother's smoking status. The present study gives supporting evidence that dietary nitrites, from which N-nitroso compounds can be formed in foods and in the human body, are associated with the development of Type 1 diabetes in man.
HLA genotypes were ascertained in 150 families with a diabetic child from the same geographical area. There was preferential zygotic assortment of the paternal HLA A1-B8 haplotype (63--65% compared with the expected 50%) in 69 diabetic families and 33 control families (pooled from elsewhere) who were informative for this haplotype. In diabetic families, the offspring also had an increased incidence of the maternal HLA A2-B15-Cw3 haplotype. Irrespective of which parent contributed the HLA A1-B8 haplotype, there was a significantly increased incidence of male children (63%) who inherited this particular haplotype. This probably explains the known excess of male diabetic children.
In a case—control design the feeding in infancy of newly diagnosed 7- to 14-year-old diabetic children (n = 426) was compared with that of age- and sex-matched non-diabetic children (n = 426) randomly selected from the Finnish population registry. All 7- to 14-year-old diabetic children diagnosed from September 1986 to the end of April 1989 from all hospitals which treat diabetic children in Finland were invited to participate in the study. Breast-feeding was initiated in almost all children, but during the birth years of this study population (1972–1982), an increase was observed in the duration of breast-feeding (whether alone or in combination with supplementary feeding) and in the age of introduction of supplementary milk feeding. The risk of Type 1 diabetes was decreased in the children who were totally breast-fed for at least 2 months (odds ratio (OR) 0.64, 95% confidence interval (CI) 0.42–0.98) or 3 months (OR 0.67, 95% CI 0.48–0.95) or exclusively breast-fed for at least 2 months (OR 0.60, 95% CI 0.41–0.89) or 3 months (OR 0.63, 95% CI 0.43–0.93). Those children who were younger than 2 months (OR 1.54, 95% CI 1.08–2.18) or 3 months (OR 1.52, 95% CI 1.11–2.08) at the time when supplementary milk feeding was begun had an increased risk of Type 1 diabetes. These associations remained significant after adjusting for the mother's education. The results suggest that early infant feeding patterns are associated with the risk of Type 1 diabetes developing at the age of 7 to 14 years.
To study the possible temporal association between primary cytomegalovirus infection and the appearance of islet cell autoantibodies or the development of insulin-dependent diabetes mellitus (IDDM) cytomegalovirus antibodies were analysed from follow-up sera of 46 initially non-diabetic siblings of diabetic children who either manifested clinical IDDM (22 siblings) or turned islet cell antibody positive (24 siblings) during the prospective observation (mean follow-up time 2.9 years). Secondly, cytomegalovirus antibodies were analysed during pregnancy in 96 mothers whose child presented with IDDM before the age of 7 years and in 96 control mothers who gave birth to a non-diabetic child. Thirdly, a case-control series including 90 newly-diagnosed young children with IDDM and their 90 control subjects was analysed. No seroconversions were found in cytomegalovirus antibodies during the follow-up of the 46 siblings indicating no temporal association with islet cell antibody seroconversion or manifestation of clinical diabetes. During the follow-up 17 (37%) siblings were constantly seronegative and 29 (63%) seropositive for cytomegalovirus IgG and there was no difference between islet cell antibody positive and negative siblings. Cytomegalovirus IgG and IgM were not different in pregnant mothers who gave birth to a subsequently diabetic child compared to control mothers, or in newly-diagnosed diabetic children compared to control children. Cytomegalovirus IgA was higher in newly-diagnosed diabetic children than in control children (p<0.005). This difference disappeared when only cytomegalovirus IgG positive individuals were analysed. No correlation was found between islet cell antibodies and cytomegalovirus antibodies in newly-diagnosed diabetic patients. The results do not support the hypothesis that primary cytomegalovirus infections could initiate the cascade leading to autoimmune destruction of the beta cells.
Animal models, most notably EMC-induced murine diabetes mellitus, have demonstrated convincingly a juvenile-type of diabetic state in the genetically susceptible host. There is evidence that one site of genetic control in mice relates to differences in viral receptors on the surface of the beta cells that may confer susceptibility or resistance to EMC-virus-induced diabetes. The availability of β cell tissue culture techniques, assuming that such cultivation does not render β cells nonspecifically susceptible to any virus, has potential in screening for possible diabetogenic viruses in man. Animal models of virus-induced diabetes in primates could offer the possibility to study carefully the evolution of diabetic complications. Certain animal models and human studies are consistent with an autoimmune component playing a part in the genesis of JIDM. A viral infection and an autoimmune process are not incompatible with each other: A virus can trigger an autoimmune event and an autoimmune phenomenon can heighten susceptibility to a a diabetogenic virus. The pathologic entity of insulitis in JIDM is compatible with a viral and/or autoimmune process being operative etiologically. Genetic studies of twins concordant and discordant for JIDM suggest an environmental component, such as a viral infection, in the etiology of this type of diabetes. HLA associations have been established with JIDM, but the nature of the gene product(s) of specific HLA haplotypes is not known. Pancreatic islet cell antibodies (ICAb) are associated with JIDM. Whether ICAb are formed in response to a viral or autoimmune process, or, whether they actually play a role in the etiology of JIDM is unclear. Epidemiologic investigations reveal a relationship of JIDM with seasonal patterns and, in certain studies, an association with increased titers of neutralizing antibodies to Coxsackie-B viruses. While a seasonal variation in JIDM can be explained by a concomitant seasonal variation in viral infections, other interpretations (such as dietary and exercise alterations) are also possible. Although these findings, taken together, suggest a role for certain viruses and other environmental factors (drugs, toxins, foods) in the pathogenesis of JIDM, further studies will be necessary to establish unequivocally such relationships and to define the specific mechanisms that are operative.
EURODIAB ACE is a collaborative European study that was set up to assess incidence of childhood insulin-dependent diabetes mellitus (IDDM) in Europe, test the proposal of a south-north gradient, and to gather information to determine the causes and pathogenesis of the disease. Here, the basic epidemiological results are reported. Newly diagnosed cases of IDDM in children aged up to 15 years were identified prospectively in twenty-four geographically well-defined study regions in Europe and Israel (a total of 16.8 million children) during 1989 and 1990. 3060 cases were identified with estimated ascertainment rates exceeding 90% in all study regions. Age-standardised and sex-standardised incidence rates varied widely, ranging from 4.6 (northern Greece) to 42.9 (two regions in Finland) cases per 100,000 per year. Rates in southern Europe were generally higher than previously assumed, and there was an unexpectedly high incidence in Sardinia, which had the second highest rate (30.2 cases per 100,000 per year) recorded in Europe. Eastern European regions had generally low rates. The collaborative network now established provides a framework for further studies to examine the complex interaction between genetic and environmental factors in the cause and pathogenesis of IDDM.
Through use of primary and secondary data sources for registration and validation, the incidence and prevalence of Type 1 (insulin-dependent) diabetes mellitus in children aged 0-14 years in Iceland has been completely ascertained for the years 1970-1989. The age-adjusted mean annual incidence per 100,000 for the 20-year period was 9.4 (95% confidence interval 7.8-11.3); similar for boys (9.9; 7.7-12.7) and girls (8.8; 6.7-11.5). Between 1970-1979 the incidence was 8.0 (6.0-10.6) and between 1980-1989 it was comparable at 10.8 (8.4-13.8) (p greater than 0.10). By Poisson regression analysis the variation in incidence was related to age at diagnosis (p less than 0.001), while a linear trend for calendar year at diagnosis did not reach statistical significance (p = 0.07). A quadratic curve, however, better described the temporal variation in incidence (p less than 0.05). The total prevalence per 1,000 by the end of 1979 and 1989 was similar, 0.45 (0.30-0.65) and 0.57 (0.40-0.79), respectively. In conclusion, this study confirms that both the incidence and prevalence of childhood Type 1 diabetes in Iceland are low compared to the other Nordic countries. The findings may suggest a causative role for environmental factors that are not related to latitude or ambient temperature.