ArticlePDF Available

Variation and trends in incidence of childhood diabetes in Europe

Authors:

Abstract

Background To study the epidemiology of childhood-onset type 1 insulin-dependent diabetes in Europe, the EURODIAB collaborative group established in 1988 prospective geographically-defined registers of new cases diagnosed under 15 years of age. This report is based on 16 362 cases registered during the period 1989-94 by 44 centres representing most European countries and Israel and covering a population of about 28 million children. Methods Multiple sources of ascertainment were used in most centres to Validate the completeness of registration by the capture-recapture method. Trends in incidence during the period were analysed by Poisson regression, the data from centres within each country being pooled. Findings The standardised average annual incidence rate during the period 1989-94 ranged from 3.2 cases per 100 000 per year in the Former Yugoslav Republic of Macedonia to 40.2 cases per 100 000 per year in two regions of Finland. By pooling over all centres, the annual rate of increase in incidence was 3.4% (95% CI 2.5-4.4%), but in some central European countries it was more rapid than this. Pooled over centres and sexes, the rates of increase were 6.3% (4.1-8.5%) for children aged 0-4 years, 3.1% (1.5-4.8%) for 5-9 years, and 2.4% (1.0-3.8%) for 10-14 years. Interpretation The results confirm a very wide range of incidence rates within Europe and show that the increase in incidence during the period varied from country to country. The rapid rate of increase in children aged under 5 years is of particular concern.
ARTICLES
THE LANCET • Vol 355 • March 11, 2000 873
Summary
Background To study the epidemiology of childhood-onset
type 1 insulin-dependent diabetes in Europe, the EURODIAB
collaborative group established in 1988 prospective
geographically-defined registers of new cases diagnosed
under 15 years of age. This report is based on 16 362 cases
registered during the period 1989–94 by 44 centres
representing most European countries and Israel and
covering a population of about 28 million children.
Methods Multiple sources of ascertainment were used in
most centres to validate the completeness of registration by
the capture-recapture method. Trends in incidence during the
period were analysed by Poisson regression, the data from
centres within each country being pooled.
Findings The standardised average annual incidence rate
during the period 1989–94 ranged from 3·2 cases per
100 000 per year in the Former Yugoslav Republic of
Macedonia to 40·2 cases per 100 000 per year in two
regions of Finland. By pooling over all centres, the annual
rate of increase in incidence was 3·4% (95% CI 2·5–4·4%),
but in some central European countries it was more rapid
than this. Pooled over centres and sexes, the rates of
increase were 6·3% (4·1–8·5%) for children aged 0–4 years,
3·1% (1·5–4·8%) for 5–9 years, and 2·4% (1·0–3·8%) for
10–14 years.
Interpretation The results confirm a very wide range of
incidence rates within Europe and show that the increase in
incidence during the period varied from country to country.
The rapid rate of increase in children aged under 5 years is of
particular concern.
Lancet 2000; 355: 873–876
Introduction
Type 1 insulin-dependent diabetes is generally believed to
be due to an immune destruction of pancreatic cells in
genetically susceptible individuals exposed to
environmental risk factors. There has been a rapid
increase in the incidence of type 1 diabetes in many
European countries in the past few decades.1–4 This
increase, in conjunction with the lack of complete
concordance in monozygotic twin pairs,5–7 points to the
importance of environmental factors.
Useful clues about these environmental factors may be
obtained by studying geographical variation in incidence
in relation to the characteristics of different countries.
Fundamental to such analyses are high-quality incidence
data uniformly collected via a standard protocol from
population-based registers. We have previously reported
data for the years 1989–90 from 26 registries in the
EURODIAB Study Group, and have established that
there is a wide range of incidence rates within Europe.8
The network has since expanded to include 44 registries,
with representation from most European countries and
Israel. Incidence data for the 6-year period 1989–94 are
presented in this paper.
Differences in the trends in incidence between
countries or age-groups may also provide important clues
about environmental factors. Some recent reports suggest
a higher rate of increase among children under 5 years of
age compared with the age groups 5–9 and 10–14 years,9,10
which suggests that the environmental factors responsible
for the increase may operate early in life. We address this
issue in our analysis of the EURODIAB registry data.
Methods
The establishment of the EURODIAB collaborative group of
childhood diabetes registers has been described in detail.8,11
Briefly, in 1988, prospective registers of new cases of insulin-
dependent type 1 diabetes mellitus among children aged under
15 years were established in 26 geographically-defined centres in
Europe and Israel. Type 1 diabetes was defined on the basis of a
clinical diagnosis of idiopathic diabetes made by a physician.
Cases secondary to other conditions (having cystic fibrosis or
high-dose steroid treatment) were excluded. Date of onset was
taken as the date of the first insulin injection. Anonymous data
were submitted to a central coordinating office in Odense,
Denmark, for data processing and analysis.
Other centres whose registries met the same quality criteria
have since joined the group, which now comprises 44 centres.
Many of the new participants are from central and eastern
Europe, and most European countries are now represented. For
the analysis of trends the 40 centres with data for all 6 years in the
study period were grouped into countries.
Capture-recapture methodology,12 which assumes the
availability of independent primary and secondary sources of
ascertainment, was used when possible to estimate the
completeless of registration. In most centres the primary source
of ascertainment was through hospital records or notifications by
paediatricians and family doctors. Secondary sources varied,
being dependent on local circumstances, but included social
insurance schemes, diabetes associations, and prescription data.
Annual estimates of the population resident in each centre’s
geographically-defined area were used as denominators for the
Variation and trends in incidence of childhood diabetes in Europe
EURODIAB ACE Study Group*
*Members of group given at the end of article
Correspondence to: Dr C C Patterson, Department of Epidemiology
and Public Health, Queen’s University of Belfast, Royal Victoria
Hospital, Grosvenor Road, Belfast BT12 6BJ, UK
(e-mail: c.patterson@qub.ac.uk)
calculation of rates. Age/sex standardised incidence rates were
obtained by the direct method with a standard population
consisting of equal numbers of children in each of six subgroups
defined by age group (0–4, 5–9, and 10–14 years) and sex.
Poisson-regression models were used to study differences in
incidence rate between countries and to investigate the trends in
incidence rate. Models with terms for sex, age group (0–4, 5–9,
and 10–14 years), country, and calendar year were fitted. These
models take account of possible differences in age/sex structure in
each country’s population and permit testing for any linear trend
in incidence rate within a country. Further models incorporating
interaction terms were used to test for differences in the linear
trends between countries, between sexes, and between age
groups. Likelihood-ratio 2tests were used to compare the fit of
nested models and to provide a test of significance for the last
term added to the model. Models were fitted by the SAS
GENMOD procedure. (SAS system for Windows, release 6.12,
SAS Institute Inc, North Carolina, USA.)
Results
Average annual incidence rates
Table 1 summarises registration for each centre—the
number of cases, the standardised average annual
incidence rate, and the estimated completeness of
ascertainment. Among the 40 centres validating
ascertainment, all but four achieved completeness of
ascertainment of over 95%. The estimates for some
centres exclude the early years of the study period, but
estimates for the years 1989–90 have already been
published.8The standardised rates varied from 3·2 cases
per 100 000 per year in the Former Yugoslav Republic of
ARTICLES
874 THE LANCET • Vol 355 • March 11, 2000
Country Region Number Standardised Completeness of
of incidence rate ascertainment
cases (95% CI) per (period)
100 000*
Austria Whole nation 753 9·1 (8·5–9·8) 99·7% (1991–94)
Belgium Antwerp 112 11·6 (9·4–13·7) 98·7% (1989–94)
Bulgaria Western 303 9·6 (8·5–10·7) 99·9% (1989–94)
Eastern 218 6·8 (5·9–7·7) 99·9% (1989–94)
Croatia Zagreb 83 6·8 (5·3–8·3) 100% (1989–94)
Czech Whole nation 1144 8·9 (8·3–9·4) 100% (1989–94)
Republic
Denmark Four counties 221 16·0 (13·9–18·1) 99·2% (1991–94)
Estonia Whole nation 206 10·3 (8·9–11·7) 100% (1989–94)
Finland Two regions 425 40·2 (36·4–44·1) 100% (1992–93)
France Four regions 837 8·3 (7·8–8·9) 99·0% (1991–94)
Germany Düsseldorf111 14·0 (11·4–16·6) 92·9% (1993–94)
Baden-Württemberg 1101 11·3 (10·6–12·0) 96·5% (1989–94)
Greece Attica 333 9·5 (8·5–10·5) 100% (1993–94)
Five northern 49 6·2 (4·5–8·0) 100% (1989–94)
regions
Hungary 18 counties 822 8·9 (8·2–9·5) 99·6% (1989–94)
Iceland Whole nation 52 13·5 (9·8–17·2) 100% (1989–94)
IsraelWhole nation 433 5·9 (5·3–6·4) 100% (1993)
Italy Lombardia 530 7·0 (6·4–7·6) NSSA
Lazio 396 8·1 (7·3–8·9) 100% (1993–94)
Sardinia 675 36·6 (33·9–39·4) 85·2% (1991–94)
Eastern Sicily 150 11·4 (9·5–13·2) 98·3% (1991–94)
Latvia Whole nation 221 6·6 (5·8–7·5) 99·7% (1993–94)
Lithuania Whole nation 368 7·4 (6·6–8·1) 100% (1989–94)
Luxemboug Whole nation 49 12·1 (8·7–15·5) 100% (1993–94)
Macedonia Whole nation 93 3·2 (2·5–3·8) 100% (1989–94)
Netherlands Five regions 421 13·0 (11·8–14·3) 95·9% (1991–94)
Norway Eight counties 491 21·2 (19·3–23·1) 99·9% (1992–94)
Poland Eight western 542 6·7 (6·2–7·3) 100% (1991–94)
provinces
Three cities 312 6·1 (5·4–6·8) 100% (1989–94)
Gliwice 316 5·4 (4·8–6·0) NSSA
Bialystok§ 31 5·5 (3·5–7·4) 100% (1994)
Portugal Madeira 24 6·9 (4·1–9·6) 100% (1993–94)
Portalegre 25 19·0 (11·5–26·5) 94·0% (1989–94)
Algarve 51 13·6 (9·8–17·3) 85·1% (1991–94)
Romania Bucharest 138 5·0 (4·1–5·8) 100% (1989–94)
Slovakia Whole nation 656 8·4 (7·7–9·0) 100% (1989–94)
Slovenia Whole nation 186 7·6 (6·5–8·7) 100% (1992–94)
Spain Catalonia 839 12·3 (11·4–13·1) 98·4% (1991–94)
Sweden Stockholm county 451 25·8 (23·4–28·2) 100% (1993–94)
Switzerland|| Whole nation 353 7·9 (7·1–8·7) NSSA
United Northern Ireland 462 19·6 (17·8–21·4) 98·6% (1989–94)
Kingdom Oxford 542 17·6 (16·1–19·1) NSSA
Leicester 169 15·9 (13·5–18·3) 100% (1989–93)
Leeds 668 15·7 (14·5–16·9) 97·6% (1989–94)
NSSA=no secondary source of ascertainment. *Standardised for age and sex.
1993–94 only. 1989–93 only. §1994 only. ||1991–94 only.
Table 1: Summary registration information for 44 EURODIAB
centres
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
Boys
Girls
Rate per 100000
Year
1989 1990 1991 1992 1993 1994
Age
10–14
5–9
0–4
Trends in childhood diabetes incidence in Europe during
1989–94 by age group and sex
Macedonia to 40·2 cases per 100 000 in the two regions
of Finland. Incidence rates were high in northern and
north-western Europe and low in central, southern, and
eastern Europe (figure 1). However, Sardinia was a
notable exception to this pattern, with a much higher rate
than any neighbouring region.
Trends in incidence
Data for each of the 6 years of the study period were
available for all but four of the centres (Germany-
Düsseldorf, Israel, Poland-Bialystok, and Switzerland).
Poisson-regression models were fitted separately for each
country, after data for the centres within each country
were pooled. Sardinia, however, was retained as a centre
on its own because the incidence there was much higher
than for other Italian centres. There was evidence of a
significant linear trend in nine of the 29 analyses, all but
one showing an increase in incidence rate (table 2). For
one country there was evidence of departure from linear
trend (p=0·04), but this was no more than might be
expected by chance and the annual rates for the country
in question showed no consistent pattern. Some central
and eastern European countries had rates that increased
rapidly (table 2).
The results of fitting Poisson-regression models to the
data from all countries simultaneously are summarised in
table 3. Preliminary model fitting confirmed that there
were significant differences in incidence rate between
countries and suggested that both sex and age effects
differed from country to country. A base model
incorporating terms for age, sex, and country was
therefore fitted to allow for differences in age/sex specific
rates from country to country. Models that specified
different patterns of linear trend were then obtained by
adding terms to the base model. The test of the overall
trend in incidence rate shown in the second line of table 3
is highly significant, with a risk ratio estimate of 1·034
(95% CI 1·025–1·044), which indicates on average a
3·4% (2·5–4·4%) increase in incidence per year. Table 3
indicates a significant difference in the trends between
countries (line 3) and a difference in the trends between
age groups (line 4) with no evidence of a difference in
trends between boys and girls (line 5) and no evidence
that the difference in trends between age-groups varied
between boys and girls (line 6). The age-specific annual
rates pooled across countries are displayed for boys and
girls in figure 2 with a logarithmic vertical scale.
Estimates of the rates of increase in the three age groups
pooled over countries and sexes were 6·3% (4·1–8·5%)
for children aged 0–4 years, 3·1% (1·5–4·8%) for 5–9
years, and 2·4% (1·0–3·8%) for 10–14 years, which
shows that in relative terms the highest rates of increase
occurred in the youngest age group.
Discussion
This multicentre study shows a greater than 10-fold
range in incidence rate of childhood diabetes in Europe.
Such variation seems unlikely to be explained by genetic
differences, since Europeans (except for some outlying
populations) are more homogeneous compared with the
indigenous populations of other continents.13 Although
the independence of primary and secondary sources of
ascertainment cannot easily be verified our assessment of
completeness is more thorough than in most previous
studies of international variations in incidence, and
underascertainment is unlikely to be a major factor in
explaining the incidence variation described here.
The rapid increase in incidence is not readily explained
by shifts in the frequency of susceptibility genes, and
change in environmental factors is a more plausible
explanation. A homogeneous pattern of increasing
incidence across age groups would suggest that similar
environmental exposures operate in each age group.
However, our observation of a greater relative increase in
incidence in those under 5 years supports the importance
of exposures operating early in life.14,15 The nature of
these exposures is not yet clear, but increased perinatal
infections16,17 or a rapid growth rate in early life18,19 could
be contributing. 6 years is too short a period to
investigate age-group differences in incidence trends
ARTICLES
THE LANCET • Vol 355 • March 11, 2000 875
Country Risk ratio (95% CI) p
per year
Austria 1·08 (1·03–1·12) 0·0005
Belgium 1·10 (0·99–1·23) 0·09
Bulgaria (two centres) 1·05 (1·00–1·10) 0·06
Croatia 1·03 (0·91–1·17) 0·68
Czech Republic 1·08 (1·04–1·12) 0·0001
Denmark 0·91 (0·84–0·99) 0·02
Estonia 1·04 (0·96–1·12) 0·36
Finland 0·99 (0·93–1·04) 0·62
France 1·04 (1·00–1·09) 0·04
Germany 1·07 (1·03–1·10) 0·0004
Greece (two centres) 0·98 (0·93–1·04) 0·57
Hungary 1·07 (1·03–1·12) 0·0006
Iceland 0·99 (0·84–1·16) 0·88
Italy (three centres) 1·01 (0·97–1·04) 0·69
Italy (Sardinia) 1·01 (0·96–1·05) 0·78
Latvia 0·98 (0·90–1·05) 0·53
Lithuania 1·00 (0·94–1·06) 0·91
Luxembourg 1·04 (0·88–1·22) 0·68
Macedonia 1·01 (0·90–1·14) 0·84
Netherlands 1·02 (0·97–1·08) 0·47
Norway 0·99 (0·94–1·04) 0·67
Poland (three centres) 1·06 (1·03–1·10) 0·0006
Portugal (three centres) 1·13 (1·01–1·27) 0·04
Romania 1·03 (0·93–1·14) 0·57
Slovakia 1·09 (1·04–1·14) 0·0003
Slovenia 1·01 (0·93–1·10) 0·86
Spain 1·00 (0·96–1·04) 0·96
Sweden 1·03 (0·98–1·09) 0·25
United Kingdom (four centres) 1·02 (0·99–1·05) 0·14
Account has been taken of changes in age/sex structure of population.
Table 2: Summary of Poisson-regression analyses showing the
incidence trend in each country during 1989–94
Model terms* Goodness of fit Likelihood ratio test for last term
2
Degrees p
2
Degrees p
of freedom of freedom
1 Base model 980·59 870 0·005 · · · · · ·
2 Base model+year 930·54 869 0·07 50·1 1 <0·001
3 Base model+year+(yearcountry) 875·00 841 0·20 55·5 28 0·001
4 Base model+year+(yearcountry)+(yearage) 865·86 839 0·25 9·14 2 0·01
5 Base model+year+(yearcountry)+(yearage)+(yearsex) 864·49 838 0·26 1·37 1 0·24
6 Base model+year+(yearcountry)+(yearage)+(yearsex)+(yearagesex) 859·91 836 0·28 4·58 2 0·10
*Base model=constant+age+sex+agesex+country+agecountry+sexcountry+agesexcountry. Age: terms for age groups 0–4, 5–9, and 10–14 years; Sex: terms for sex; Country:
terms for country; Year: terms for linear trend across the 6 years; Agesex: terms for the interaction between age-group and sex.
Table 3: Summary of Poisson-regression analyses of incidence trends for data from 40 centres grouped into 29 countries
within each centre. Our modelling compared age-group
differences in incidence trends pooled over centres
instead, which took account of the different trends in the
various centres.
It is tempting to link the rapid increase in incidence
rates in some central and eastern European countries
with recent political changes. However, the long
prediabetic phase thought to be involved in the
pathogenesis of this disease20 suggests that the factors
responsible for these rapid increases may have operated
earlier. It will be important to continue monitoring the
incidence rate trends in these countries and to compare
them with those in Romania and the republics of former
Yugoslavia where increases in incidence rate are not yet
apparent.
Project co-ordination: A Green, Department of Epidemiology and Social
Medicine, University of Aarhus, Denmark; G Brutti, EURODIAB ACE
Co-ordinating Office, Odense University Hospital, Denmark.
Writing committee: C C Patterson, Department of Epidemiology and
Public Health, Queen’s University Belfast, Northern Ireland;
G Dahlquist, Department of Clinical Science, Pediatrics, University of
Umeå, Sweden; G Soltész, Department of Pediatrics, University of Pécs,
Hungary; A Green, Department of Epidemiology and Social Medicine,
University of Aarhus, Denmark.
Study centre leaders: Austria—E Schober, Department of Paediatrics,
University of Vienna. Belgium—I Weets, C Vandevalle, F Gorus,
M Coeckelberghs, M Du Caju, Belgian Diabetes Registry, Brussels.
Bulgaria (two centres)—V Christov, Clinic of Endocrinology, University
Alexandrov Hospital, Sofia. V Tzaneva, V Iotova, Department of
Pediatrics, Clinic of Endocrinology, Medical University, Varna. Croatia
G Roglic, Vuk Vrhovac Institute, Zagreb. Czech Republic—J Vavrinec,
Second Clinic of Pediatrics, Charles University, Prague. Denmark
B S Olsen, A J Svendsen, J Kreutzfeldt, E Lund, Department of
Pediatrics, KAS-Glostrup. Estonia—T Poodar, Hospital of
Endocrinology, Tartu. Finland—J Tuomilehto, M Karvonen, Diabetes
and Genetic Epidemiology Unit, National Public Health Institute,
Helsinki. France—C Levy-Marchal, P Czernichow, J Doutreix, INSERM
U457, Service d’Endocrinologie et Diabétologie Pédiatrique, Hôpital
Robert Debré, Paris. Germany (two centres)—G Giani, Diabetes Research
Institute, Düsseldorf; A Neu, Tübingen Kinderklinik-Sektion Päd
Endokrinol, Eberhard-Karls-Universität. Greece (two centres)—C
Bartsocas, K Kassiou, C Dacou-Voutetaki, A C Kafourou, A Al-Qadreh,
C Karagianni, Department of Pediatrics, National University of Athens;
N Papazoglou, General Hospital Agios Pavlos, Thessaloniki. Hungary
G Soltész, Department of Paediatrics, University of Pécs. Iceland
A V Thorsson, Department of Pediatrics, University of Iceland, Reykjavik.
Israel—Z Laron, O Gordon, Y Albag, I Shamis, Paediatric Endocrinology
and Diabetes Reseach Unit, Petah Tikva. Italy (four centres)—
G Chiumello, Clinica Pediatrica III, Istituto Scientifico H San Raffaele,
Milano; P Pozzilli, N Visalli, L Sebastiani, G Marietti, R Buzzetti,
Universitá Campus Biomedico, Rome; M Songini, A Casu, A Marinaro,
R Ricciardi, M A Zedda, A Milia, Department of Internal Medicine,
Hospital San Michele, Cagliari; F Purrello, M Arpi, G Fichera,
M Mancuso, C Lucenti, Department of Endocrinology, Ospedale
Garibaldi, Catania. Latvia—G Brigis, Public Health and Epidemiology,
Latvian Academy of Medicine, Riga. Lithuania—B Urbonaité, Institute of
Endocrinology, Kaunas Medical Academy. Luxembourg—C De Beaufort,
Clinique Pediatrique de Luxembourg. Macedonia—M Kocova, Pediatric
Clinic, Medical Faculty, University of Skopje. Netherlands—M Reeser,
Juliana Children’s Hospital, The Hague. Norway—G Joner, Department
of Community Health, Folkehelsa-Epidemiology, Oslo. Poland (four
centres)—D Woznicka, Pediatrics, Endocrinology and Diabetes,
University of Medical Sciences, Poznan; Z Szybinski, Department of
Endocrinology, Jagiellonian University, Krakow; P Jarosz-Chobot,
Department of Children’s Endocrinology, Silesian School of Medicine,
Katowice; I Kinalska, Department of Endocrinology, University Hospital,
Bialystok. Portugal (three centres)—S Abreu, Unidade de Endocrinologia,
Centro Hospitalar do Funchal, Madeira; C Menezes, Servico de Medicina
Interna, Hospital Distrital de Portalegre; E A Pina, Servico de Medicina,
Hospital Distrital de Faro. Romania—C Ionescu-Tirgoviste, Department
of Nutrition and Metabolic Disease, University of Bucharest. Slovakia
D Michalková, P Hlava, M Mikuleck´y, J ˘
Cernay, First Pediatric Clinic,
University Komensky, Bratislava. Slovenia—C Krzisnik, T Battelino,
N Bratina-Ursic, Endocrine, Diabetes and Metabolic Diseases, University
Medical Centre, Ljubljana. Spain—A Goday, Servicio d’Endocrinologia,
Hospital del Mar, Barcelona. Sweden—G Dahlquist, Department of
Clinical Science, Pediatrics, University of Umeå. Switzerland—E Schönle,
Univ-Kinderklinik Eleonoren-Stiftung, Kinderspital Zürich. UK (four
centres)—C Patterson, R Greenlees, D Carson, D Hadden, Queen’s
University Belfast, Royal Victoria Hospital, Belfast; P Bingley,
Department of Medicine, Southmead Hospital, University of Bristol;
N Raymond, Department of Epidemiology and Public Health, University
of Leicester; P McKinney, H Bodansky, C Stephenson, Institute of
Epidemiology and Health Services Research, University of Leeds.
Acknowledgments
This study has been supported in part by the European Community
Concerted Action Programme (grants BMH1-CT92-0043 and
BMH4-CT96-0577).
References
1 Bingley PJ, Gale EAM. Rising incidence of IDDM in Europe. Diabetes
Care 1989; 12: 289–95.
2 Diabetes Epidemiology Research International Group. Secular trends
in incidence of childhood IDDM in 10 countries. Diabetes 1990; 39:
858–64.
3 Karvonen M, Tuomilehto J, Libman I, LaPorte R. A review of the
recent epidemiological data on the worldwide incidence of type 1
(insulin-dependent) diabetes mellitus. Diabetologia 1993; 36: 883–92.
4 Green A, Sjolie AK, Eshoj O. The epidemiology of diabetes mellitus.
In: Pickup JC, Williams G, eds. Textbook of diabetes. Oxford:
Blackwell Science, 1997.
5 Leslie RDG, Lo SSS, Hawa M, et al. Lessons on the etiology of
insulin-dependent diabetes from twin studies. In: Levy-Marchal C,
Czernichow P, eds. Epidemiology and etiology of insulin-dependent
diabetes in the young. Basle: Karger, 1992.
6 Kaprio J, Tuomilehto J, Koskenvuo M, et al. Concordance for type 1
(insulin-dependent) and type 2 (non-insulin dependent) diabetes
mellitus in a population-based cohort of twins in Finland. Diabetologia
1992; 35: 1060–67.
7 Kyvik KO, Green A, Beck-Nielsen H. Concordance rates of insulin
dependent diabetes mellitus: a population based study of young
Danish twins. BMJ 1995; 311: 913–17.
8 Green A, Gale EAM, Patterson CC, for the EURODIAB ACE Study
Group. Incidence of childhood-onset insulin-dependent diabetes
mellitus: the EURODIAB ACE study. Lancet 1992; 339: 905–09.
9 Tuomilheto J, Virtala E, Karvonen M, et al. Increase in incidence of
insulin-dependent diabetes mellitus among children in Finland. Int J
Epidemiol 1995; 24: 984–92.
10 Gardner SG, Bingley PJ, Sawtell PA, Weeks S, Gale EAM. Rising
incidence of insulin dependent diabetes in children aged under 5 years
in the Oxford region: time trend analysis. BMJ 1997; 315: 713–17.
11 Levy-Marchal C, Patterson C, Green A, on behalf of the EURODIAB
ACE Study Group. Variation by age-group and seasonality at
diagnosis of childhood IDDM in Europe. Diabetologia 1995; 38:
823–30.
12 Bishop YMM, Fienberg SE, Holland PW. Estimating the size of a
closed population. Discrete multivariate analysis: theory and practice.
Cambridge, Massachusetts, MIT Press, 1975.
13 Cavalli-Sforza LL, Piazza A. Human genomic diversity in Europe: a
summary of recent research and prospects for the future. Eur J Hum
Genet 1993; 1: 3–18.
14 Leslie RD, Elliott RB. Early environmental events as a cause of
IDDM: evidence and implications. Diabetes 1994; 43: 843–50.
15 Dahlquist GG. Viruses and other perinatal exposures as initiating
events for beta-cell destruction. Ann Med 1997; 29: 413–17.
16 Dahlquist GG, Ivarsson S, Lindberg B, Forsgren M. Maternal
enteroviral infection during pregnancy as a risk factor for childhood
IDDM. A population-based case-control study. Diabetes 1995; 44:
408–13.
17 Hyöty H, Hiltunen M, Knip M, et al. A prospective study of the role
of Coxsackie B and other enterovirus infections in the pathogenesis of
IDDM. Diabetes 1995; 44: 652–57.
18 Blom L, Persson LA, Dahlquist G. A high linear growth is associated
with an increased risk of childhood diabetes mellitus. Diabetologia
1992; 35: 528–33.
19 Johansson C, Samuelsson U, Ludvigsson J. A high weight gain early in
life is associated with an increased risk of type 1 (insulin-dependent)
diabetes mellitus. Diabetologia 1994; 37: 91–94.
20 Gorsuch AN, Spencer KM, Lister J, et al. Evidence for a long
prediabetic period in type I (insulin-dependent) diabetes mellitus.
Lancet 1981; ii: 1363–65.
ARTICLES
876 THE LANCET • Vol 355 • March 11, 2000
... Type 1 diabetes is major diabetes among youth and it effects the youth of age group 20s worldwide (Group, 2006b) , (Vandewalle et al., 1997) , (Thunander et al., 2008). The prevalence rate peaks from birth to age group of 10-14 years during the onset of puberty (Dabelea et al., 2007), (Group, 2000). The incidence rate of diabetes type 1 patients is elevated in people of age group 0-4 years in Europe (Group, 2000). ...
... The prevalence rate peaks from birth to age group of 10-14 years during the onset of puberty (Dabelea et al., 2007), (Group, 2000). The incidence rate of diabetes type 1 patients is elevated in people of age group 0-4 years in Europe (Group, 2000). The incidence rate dwindles after passing the puberty age and tends to sustain in the age group 15-29 years. ...
Article
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.
... In Lombardy, the incidence of T1D in pediatric age has increased by 128% in 31 years from 7/100¢000 in 1989 to 16/100¢000 in 2020 (about 4%/year) in a non-linear manner with an increase in the years 1990À2000 and then stable oscillating around 15/100¢000 in the last 15 years [18,19]. Especially young children 0À4 years seemed to be slightly more affected, similarly to our findings [18]. ...
... In Lombardy, the incidence of T1D in pediatric age has increased by 128% in 31 years from 7/100¢000 in 1989 to 16/100¢000 in 2020 (about 4%/year) in a non-linear manner with an increase in the years 1990À2000 and then stable oscillating around 15/100¢000 in the last 15 years [18,19]. Especially young children 0À4 years seemed to be slightly more affected, similarly to our findings [18]. Therefore, in the next few years we are expecting to diagnose and manage an increasing number of children and adolescents affected by T1D. ...
Article
Background: The Italian Lombardy region has been the epicenter of COVID-19 since February 2020. This study analyses the epidemiology of pediatric type 1 diabetes (T1D) onset during the first two pandemic waves and three previous years. Methods: All the 13 pediatric diabetes centers in Lombardy prospectively evaluated charts of children at T1D onset (0-17 years), during year 2020. After calculating the annual incidence, the data were compared with those of the 3 previous years, using generalized linear models, adjusted for age and sex. Monthly T1D new onsets and diabetic ketoacidosis (DKA) were investigated yearly from 2017 to 2020. Data were extracted from outpatients charts of the pediatric diabetes centers and from the database of the national institute of statistics. Findings: The estimated incidence proportion of T1D was 16/100,000 in 2020, compared to 14, 11 and 12 in 2019, 2018 and 2017, respectively. When adjusting for age and gender, the incidence was significantly lower in 2018 and 2017 compared to 2020 (adjusted incidence ratio: 0.73 and 0.77 respectively, with 95% CI: 0.63 to 0.84, and 0.67 to 0.83; p = 0.002 and p = 0.01), but no difference was found between the years 2020 and 2019. A reduction trend in the percentage of T1D diagnosis during the first wave (March-April) over the total year diagnoses was observed compared to previous years (11.7% in 2020, 17.7% in 2019, 14.1% in 2018 and 14.4% 2017). No difference was observed during the second wave (October-December) (32.8% in 2020, 33.8% in 2019, 34% in 2018, 30.7% in 2017). The proportion of DKA over the total T1D diagnoses during the second wave had higher trend than the first one (41.7% vs 33.3%), while severe DKA over the total DKA appeared higher during the first wave (60% vs 37.1%). Interpretation:The study suggests an increase in the incidence of pediatric T1D in Lombardy throughout thepastfive years. Pandemic waves may have affected the clinical presentation at onset. Funding:None.
... • 2.3% in Colorado [28]. • 3.4% in EURODIAB [29]. ...
Article
Full-text available
Purpose of Review Epidemiological research on type 1 diabetes (T1D) has traditionally focussed on the paediatric age group, but recent data in adults has confirmed it to be a disease of all ages with a wide clinical spectrum. We review the epidemiology and clinical features of T1D across the lifespan. Recent findings While the peak incidence of T1D is still in early adolescence, T1D is now diagnosed more commonly in adulthood than childhood due to increasing recognition of adult-onset T1D and the length of the adult lifespan. It still follows the known geographic variations in incidence, being highest in Northern Europe and lowest in Asia. The onset of T1D in adulthood is usually less acute than in childhood and confers a lower, although still substantial, risk of complications and early mortality. Interventions to delay T1D onset are emerging and screening for those at risk at birth is increasingly available. Summary Type 1 diabetes can develop at any age and may not present with ketosis or an immediate insulin requirement in adults. Macro- and microvascular complications are the greatest cause of excess morbidity and mortality in this population.
... The main complications of diabetes, osteopenia, and bone microstructural changes increase the possibilities of fractures and friability and decrease bone strength (4). A meta-analysis of 16 studies reported no difference in lumbar spine bone marrow density (BMD) between adults with type I diabetes (T1D) and healthy subjects, controlling for age and gender effect (5). The associations of glycemic control and diabetes duration with BMD reported in many studies have been contradictory, but microvascular complications (retinopathy, neuropathy, and nephropathy) have been found in low-BMD patients with T1D (6)(7)(8). ...
Article
Full-text available
Purpose To investigate whether the microvascular permeability of lumbar marrow and bone trabecular changes in early-stage diabetic rabbits can be quantitatively evaluated using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), quantitative computed tomography, and texture-analyzed permeability parameter K trans map of DCE-MRI. Materials and Methods This prospective study included 24 rabbits that were randomly assigned to diabetic (n = 14) and control (n = 10) groups. All rabbits underwent sagittal MRI of the lumbar region at 0, 4, 8, 12, and 16 weeks after alloxan injection. Pearson correlation coefficient was performed to determine the correlation between permeability parameter and bone mineral density (BMD). Repeated-measures ANOVA was used to analyze the changes in lumbar BMD over time in each group and the texture parameters of diabetic rabbit lumbar marrow at different time points. Mann–Whitney U rank sum test was used to compare the differences of each index between the two groups and calculate the area under the curve (AUC). Results BMD was correlated with K trans , K ep , and V e but not with V p . At weeks 0–16, the BMD of the rabbits in the diabetic and normal groups was not statistically significant, but the change in BMD showed an overall downward trend. For texture analysis, entropy, energy, and Uniformized positive pixel (UPP) parameters extracted from the K trans map showed significant differences from week 0 to 16 between the two groups. The identification ability at 8–12 weeks was higher than that at 12–16 weeks, and the AUCs were 0.734, 0.766, and 0.734, respectively (P < 0.05 for all). Conclusions The changes in BMD measured using quantitative computed tomography occurred later than those measured using bone trabecular morphometry. Texture analysis parameters based on DCE-MRI quantitative parameter Ktrans map are feasible to identify early changes in lumbar marrow structure in diabetic rabbits.
... Incidence reaches 3% of the child population. It has been estimated that approximately 96,000 new cases of diabetes are revealed annually [3,4]. Type 1 diabetes is caused by the destruction of the β-cells of the pancreas, which leads to an absolute insulin deficiency. ...
Article
Full-text available
Type 1 diabetes (DM1) is classified as an autoimmune disease. An uncontrolled response of B and T lymphocytes to the body’s own tissues develops in the absence of immune tolerance. The main aim of the study was to evaluate the effect of the duration of type 1 diabetes in children on the expression of TLR receptors and the relationship with the parameters of glycemic control in patients. As a result, we showed significant differences in the level of TLR2, TLR4 and TLR9 expression in patients with DM1 in the early stage of the disease and treated chronically compared to the healthy group. Additionally, in this study, we found that the numbers of CD19+ B cells, CD3+ CD4+, CD3+ CD8+ T cells and NK cells are different for newly diagnosed DM1 individuals, patients receiving chronic treatment and for healthy controls, indicating an important role of these cells in killing pancreatic beta cells. Moreover, higher levels of IL-10 in patients with newly diagnosed DM1 have also been found, confirming the reports found in the literature.
Article
Full-text available
Annually the number of people with diabetes is increasing. Therefore diagnosing this disease in the early stages to save human lives is a great challenge. Different strategies have been reported for the monitoring and treatment of diabetes such as:lable-free detection of abnormal biomolecules,sensors for exhaled breath anlysis, biosensor based photonic crystal structure,plasmonic chip for biomarker detection. These techniques have limitations like: high-cost, not available for clinical diagnosis, not enough sensitive and time consuming. With discovering two-dimensional 2D materials, extensive research has been conducted based on these materials. Different 2D materials like graphene and its derivations, transition metal chalcogenides TMD, boron nitride, phosphorene, nanosheets, MXene and iii-vi layered has attracted great attention for development of state of the art devices and applications. In the one hand by changing the 2D layers number, different optical absorption can be obtained. Moreover 2D materials exhibit remarkable potential in optical sensing, bio-medical detection and energy-efficient due to their outstanding physical, chemical, optical and electronic properties. Extensive novel properties of 2D materials such as, specificity, high surface to area ratio, environmental stability, rapid and time consuming performance pave the fabrication and design of devices based 2D material.
Article
Objective: The Environmental Determinants of Diabetes in the Young (TEDDY) study is uniquely capable of investigating age-specific differences associated with type 1 diabetes. Because age is a primary driver of heterogeneity in type 1 diabetes, we sought to characterize by age metabolic derangements prior to diagnosis and clinical features associated with diabetic ketoacidosis (DKA). Research design and methods: The 379 TEDDY children who developed type 1 diabetes were grouped by age at onset (0-4, 5-9, and 10-14 years; n = 142, 151, and 86, respectively) with comparisons of autoantibody profiles, HLAs, family history of diabetes, presence of DKA, symptomatology at onset, and adherence to TEDDY protocol. Time-varying analysis compared those with oral glucose tolerance test data with TEDDY children who did not progress to diabetes. Results: Increasing fasting glucose (hazard ratio [HR] 1.09 [95% CI 1.04-1.14]; P = 0.0003), stimulated glucose (HR 1.50 [1.42-1.59]; P < 0.0001), fasting insulin (HR 0.89 [0.83-0.95]; P = 0.0009), and glucose-to-insulin ratio (HR 1.29 [1.16-1.43]; P < 0.0001) were associated with risk of progression to type 1 diabetes. Younger children had fewer autoantibodies with more symptoms at diagnosis. Of 23 (6.1%) of 379 children with DKA at onset, only 1 (4.3%) of 23 had a first-degree relative (FDR) with type 1 diabetes compared with 102 (28.7%) of 356 FDR children without DKA (P = 0.008). Children with DKA were more likely to be nonadherent to study protocol (P = 0.047), with longer duration between their last TEDDY evaluation and diagnosis (median 10.2 vs. 2.0 months without DKA; P < 0.001). Conclusions: DKA at onset in TEDDY is uncommon, especially for FDRs. For those without familial risk, metabolic monitoring continues to provide a primary benefit of reduced DKA but requires regular follow-up. Clinical and laboratory features vary by age at onset, adding to the heterogeneity of type 1 diabetes.
Book
Full-text available
Chapter 1. Introduction Over the past three centuries, health of Europeans has improved spectacularly, with a doubling of life expectancy at birth. Below the surface of these changes in life expectancy lies a fascinating pattern of rising and then falling diseases. The explanation of these trends has been a battle of ideas, between those who believed rising life expectancy to be a by-product of economic development, and others who have emphasized the role of human agency, in the form of public health and medical care. This book exploits an unique dataset of long-term trends in around 40 diseases, and their variations between European countries, to identify the main drivers of secular trends in population health. Chapter 2. Long-term trends in population health This chapter reviews secular trends in mortality, causes of death, morbidity, height, and disability-free life expectancy since 1700. While mortality declined, trends in morbidity have been somewhat less favourable, so that the number of both healthy and unhealthy life-years has increased. A preview of the disease-specific changes presented in subsequent chapters shows that trends have mostly occurred in a striking pattern of ‘rise-and-fall’. These findings are then used to present a modified version of the ‘epidemiologic transition’, with three stages: a first stage in which ‘health problems of pre-industrial societies’ started to decline, a second stage in which ‘health problems of industrializing societies’ started to decline, and a third stage in which ‘health problems of affluent societies’ started to decline. Chapter 3. Understanding trends in population health As a theoretical background to the analysis of long-term changes in population health, this chapter presents an ‘ecological-evolutionary theory’ of the origins of disease, which stipulates that most diseases are caused by ‘unfavourable exchanges’ between the human organism and its external environment, often in combination with one or more ‘failures in the design’ of the organism. The crucial role of environmental factors implies that population health is very sensitive to economic, political and sociocultural conditions. This chapter then reviews long-term trends in these conditions, such as improvements in living standards, the rise of the modern state, and the advent of the ‘Enlightenment’, as well as changes in the organization and effectiveness of public health and medical care. All these factors have made demonstrable contributions to improvements in population health. Chapter 4. Health problems of pre-industrial societies In the pre-industrial period, mortality trends in Europe were characterized by frequent mortality crises, often in connection with war, famine and/or epidemics. This chapter reviews long-term trends in wars and war-related deaths, in homicide, and in the occurrence of famines, and then looks at secular trends in four diseases that often caused massive epidemics: plague, smallpox, typhus and malaria. The rise of these health problems, often in the distant past, is traced, as well as their ultimate decline. The factors involved in their decline range from better diplomacy to draining marshes, and from vaccination to the ‘civilization process’, and were often facilitated by economic, political and sociocultural change. There were striking differences between European regions in the timing of the decline of health problems of pre-industrial societies, with Northern and Western Europe taking the lead. Chapter 5. Health problems of industrializing societies Industrialization and urbanization were accompanied by a rise and then decline of many different diseases. This chapter first traces the history of a number of communicable diseases, including three intestinal infections (cholera, dysentery and typhoid), tuberculosis, syphilis, four childhood infections (scarlet fever, measles, whooping cough and diphtheria ) and two respiratory infections (pneumonia and influenza). It then traces long-term trends in maternal, infant and perinatal mortality, and in three nutrient deficiencies (pellagra, rickets and goitre), peptic ulcer and appendicitis, and lung diseases caused by occupational and environmental exposures (such as pneumoconiosis, mesothelioma and the non-specific effects of air pollution). The factors involved in the ultimate decline of these diseases were many, with an important role for public health interventions. There were striking differences between European regions in the timing of the decline of health problems of industrializing societies, with Northern and Western Europe again often taking the lead. Chapter 6. Health problems of affluent societies After the conquest of many infectious diseases and other health problems of industrializing societies, morbidity and mortality patterns in Europe became dominated by a range of chronic diseases, including ischaemic heart disease, cerebrovascular disease, diabetes mellitus, various cancers, liver cirrhosis, dementia, and depression, as well as by injuries, including road traffic injuries and suicide. More recently, a new ‘plague’ occurred in the form of AIDS. This chapter traces long-term trends in these diseases, which again often manifested themselves in a striking pattern of ‘rise-and-fall’. Among the factors involved in the ultimate decline of these diseases, improvements in the effectiveness of medical care now also played a prominent role than in the past, but economic, political and sociocultural changes were still important in the background. As in previous periods, there were striking differences between European regions in the timing of the decline of these health problems, with Northern, Western and Southern Europe taking the lead. Chapter 7. Why? This chapter returns to a more general view-point, and addresses a number of deeper ‘why’-questions. Why did European population health improve? Based on the findings of the previous chapters, an attempt is made to apportion credit to different forms of human agency, including public health and medical care. It is argued that, in the final instance, these population health improvements were caused by an increase in rational thinking. And why did some countries rush ahead or lag behind? To address this question, this chapter reviews the experience of five extraordinary countries, namely Sweden, the Netherlands, Spain, Albania and Russia. It concludes on an important role for both political and sociocultural conditions in shaping long-term trends in population health. Chapter 8. Outlook This short chapter contains a summary of the main conclusions of this book, as well as an analysis of the main risks to keeping population health at its current, high levels. Threats include increased geopolitical instability, increasing socioeconomic inequality, and global environmental changes such as climate change. We also can no longer ignore the damage that has been done to other living species, which is one of the darkest sides of human progress in the previous centuries.
Article
Full-text available
Sumario: Structural models for counted data -- Maximum likelihood estimates for complete tables -- Formal goodness of it: summary statistics and model selection -- Maximum likelihood estimation for incomplete tables -- Estimating the size of a closed population -- Models for measuring change -- Analysis of square tables: symmetry and marginal homogeneity -- Model selection and assessing closeness of fit: practical aspects -- Other methods for estimation and testing in cross-classifications -- Measures of association and agreement -- Pseudo-bayes estimates of cell probabilities -- Sampling models for discrete data -- Asymptotic methods.
Article
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.
Article
There is strong evidence that the aetiology of insulin-dependent diabetes mellitus (IDDM) is due to a complex interaction between genes and the environment and that the pathogenesis is autoimmune. In early perinatal life the immune system is induceable and exposures in this period may initiate autoimmunity. Recent findings of time and space clustering of birth dates for later diabetic cases together with the early observa- tion of a very high prevalence of diabetes in cases with rubella embryopathy suggest that foetal virus exposure may be important. Recent findings from Sweden and Finland suggest that enterovirus exposure during foetal life may initiate autoimmunity which may lead to diabetes. Other immune events, such as maternal-foetal blood group incompatibility and pre-eclampsia in the mother have also been associated with IDDM risk. Other more unspecific events in the perinatal period, such as a short gestational age, caesarean section and neonatal respiratory disease, are also indicated to increase the risk. In addition, food components such as nitrosamine components, cow's milk protein and gliadin have been proposed to initiate the slowly progressing autoimmune b-cell destruction. Most of these epidemiological findings are supported by experi- mental studies in the nonobese diabetic mice but their exact mechanisms of action are still unclear. It is concluded that new evidence is accumulating indicating that perinatal exposures may be important for the initiation of b-cell destruction. As such risk factors may be targets for primary prevention strategies further studies are urgently warranted.
Chapter
IntroductionEvidence of a Genetic Contribution to Type 1 DiabetesEvidence of Environmental DeterminantsPredicting Risk of Type 1 Diabetes: Available MarkersPredicting Risk of Type 1 Diabetes: Methodological ConsiderationsPredicting Risk of Type 1 Diabetes: A Hyphothetical ExampleConcluding RemarksReferences
Chapter
"At last, after a decade of mounting interest in log-linear and related models for the analysis of discrete multivariate data, particularly in the form of multidimensional tables, we now have a comprehensive text and general reference on the subject. Even a mediocre attempt to organize the extensive and widely scattered literature on discrete multivariate analysis would be welcome; happily, this is an excellent such effort, but a group of Harvard statisticians taht has contributed much to the field. Their book ought to serve as a basic guide to the analysis of quantitative data for years to come." -James R. Beninger, Contemporary Sociology "A welcome addition to multivariate analysis. The discussion is lucid and very leisurely, excellently illustrated with applications drawn from a wide variety of fields. A good part of the book can be understood without very specialized statistical knowledge. It is a most welcome contribution to an interesting and lively subject." -D.R. Cox, Nature "Discrete Multivariate Analysis is an ambitious attempt to present log-linear models to a broad audience. Exposition is quite discursive, and the mathematical level, except in Chapters 12 and 14, is very elementary. To illustrate possible applications, some 60 different sets of data have been gathered together from diverse fields. To aid the reader, an index of these examples has been provided. ...the book contains a wealth of material on important topics. Its numerous examples are especially valuable." -Shelby J. Haberman, The Annals of Statistics. © 2007 Springer Science+Business Media, LLC. All rights reserved.
Article
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.
Article
We studied the cumulative incidence, concordance rate and heritability for diabetes mellitus in a nationwide cohort of 13,888 Finnish twin pairs of the same sex. The twins were born before 1958 and both co-twins were alive in 1967. Data on diabetes were derived through computerized record linkage from death certificates, the National Hospital Discharge Register and the National Drug Register. Records were reviewed in order to assign a diagnostic category to the 738 diabetic patients identified. Of these patients 109 had Type 1 (insulin-dependent) diabetes, 505 Type 2 (non-insulin-dependent) diabetes, 46 gestational diabetes, 24 secondary diabetes, 38 impaired glucose tolerance and 16 remained unclassified. The cumulative incidence of diabetes was 1.4% in men and 1.3% in women aged 28-59 years and 9.3% and 7.0% in men and women aged 60 years and over, respectively. The cumulative incidence did not differ between monozygotic and dizygotic twins. The concordance rate for Type 1 diabetes was higher among monozygotic (23% probandwise and 13% pairwise) than dizygotic twins (5% probandwise and 3% pairwise). The probandwise and pairwise concordance rates for Type 2 diabetes were 34% and 20% among monozygotic twins and 16% and 9% in dizygotic twins, respectively. Heritability for Type 1 diabetes was greater than that for Type 2 where both genetic and environmental effects seemed to play a significant role.
Article
Insulin release and growth are intimately connected. The aim of the present study was to investigate height and weight in diabetic children from birth to onset of Type 1 (insulin-dependent) diabetes mellitus compared to that in referent children. Data on height and weight were collected from mailed questionnaires and from growth records obtained from the child health clinics and schools in 337 recent-onset diabetic children, 0-14 years old, and from 517 age-, sex-, and geographically matched referent children. A total of 9002 paired height and weight observations were collected. The anthropometric development of the children was expressed as standard deviation scores using the National Center for Health Statistics/Centers for Disease Control (NCHS/CDC) growth reference material. On the average, the diabetic children were consistently taller than the referent children, a finding more pronounced among the boys. The diabetic boys were significantly taller from 7 to 1 years before the clinical onset of the disease, regardless of age at onset. A similar tendency was found for the girls. When mean height from 5 to 1 years before onset was used as a possible risk factor for diabetes, a linearly increasing trend in the odds ratio was found for diabetes in boys (odds ratio = 1.0; 1.57; 2.46 for height standard deviation score values less than 0; 0-1 and greater than 1, respectively; p = 0.002 for trend). A similar, but statistically not significant, tendency was found for girls (odds ratio = 1.0; 1.44; 1.43). As regards height increment from birth similar trends in odds ratios were found.(ABSTRACT TRUNCATED AT 250 WORDS)
Article
A rising incidence of insulin-dependent diabetes mellitus (IDDM) has been reported in many northern European countries, with a rate equivalent to a doubling time of 20-30 yr in some. North American and Japanese studies report a similar trend, although they are less uniform in their findings. Although the number of genetically susceptible individuals within these populations has increased, the rapidity of the change suggests that environmental factors are responsible. If these could be identified, primary prevention might become possible.