Relative predispositional effects of HLA class II DRB1-DQB1 haplotypes and genotypes on type 1 diabetes: a meta-analysis.

The direct involvement of the human leukocyte antigen class II DR-DQ genes in type 1 diabetes (T1D) is well established, and these genes display a complex hierarchy of risk effects at the genotype and haplotype levels. We investigated, using data from 38 studies, whether the DR-DQ haplotypes and genotypes show the same relative predispositional effects across populations and ethnic groups. Significant differences in risk within a population were considered, as well as comparisons across populations using the patient/control (P/C) ratio. Within a population, the ratio of the P/C ratios for two different genotypes or haplotypes is a function only of the absolute penetrance values, allowing ranking of risk effects. Categories of consistent predisposing, intermediate ('neutral'), and protective haplotypes were identified and found to correlate with disease prevalence and the marked ethnic differences in DRB1-DQB1 frequencies. Specific effects were identified, for example for predisposing haplotypes, there was a statistically significant and consistent hierarchy for DR4 DQB1*0302s: DRB1*0405 =*0401 =*0402 > *0404 > *0403, with DRB1*0301 DQB1*0200 (DR3) being significantly less predisposing than DRB1*0402 and more than DRB1*0404. The predisposing DRB1*0401 DQB1*0302 haplotype was relatively increased compared with the protective haplotype DRB1*0401 DQB1*0301 in heterozygotes with DR3 compared with heterozygotes with DRB1*0101 DQB1*0501 (DR1). Our results show that meta-analyses and use of the P/C ratio and rankings thereof can be valuable in determining T1D risk factors at the haplotype and amino acid residue levels.

14th International HLA and Immunogenetics Workshop: report on the HLA component of type 1 diabetes.

The type 1 diabetes (T1D) component of the 13th International Histocompatibility Workshop (IHW) obtained microsatellite (msat) and human leukocyte antigen (HLA)-DR/DQ data on case/control and family samples through an international collaboration. The aim was to detect the effects of susceptibility loci on the HLA complex independent of the primary determinants in the class II region (HLA-DR/DQ). As part of the activity of the 14th International HLA and Immunogenetics Workshop (14th IHIWS), a T1D workshop was held to present analyses of the 13th IHW data and to discuss the current status of knowledge about the genetics of T1D. These data are now available online through dbMHC, a web-based resource established by the National Center for Biotechnology. Continuing work since the 13th IHW has resulted in published work showing heterogeneity of DR3 haplotypes in data sets from the 13th IHW and Human Biological Data Interchange (HBDI). In addition, we identified markers that define DRB1*1501 DQB1*0602 haplotypes conferring reduced protection from diabetes in a Swedish 13th IHW data set. Further analyses of the 13th IHW data set not only showed some significant results but also demonstrated extensive heterogeneity reminiscent of non-HLA genes. The haplotype analysis in HBDI families identified two msats with significant effects on susceptibility and statistically significant age of onset effects at class III markers that are not because of linkage disequilibrium, with class I alleles known to affect age of onset. The above studies underscore the importance of refining our understanding of susceptibility associated with genes in the HLA complex.

D6S265*15 marks a DRB1*15, DQB1*0602 haplotype associated with attenuated protection from type 1 diabetes mellitus.

AIMS/HYPOTHESIS: The HLA class II DQB1*0602 allele confers strong dominant protection against type 1 diabetes but protection is not absolute. The aim of this study was to identify markers within the HLA region that differentiate DQB1*0602 haplotypes and show different associations with disease risk. METHODS: We defined alleles at eight microsatellite markers spanning the HLA region in a case-control cohort from Sweden. RESULTS: We found that allele 15 at marker D6S265 (109 kb centromeric of HLA-A) was over-represented among patients carrying DRB1*15, DQB1*0602. A detailed haplotype analysis showed that DRB1*15, DQB1*0602 haplotypes carrying D6S265*15 have a ten-fold higher odds ratio (OR) than those carrying other alleles and thus confer reduced protection [OR D6S265*15=0.186 (95% CI 0.074, 0.472) vs OR D6S265*15-=0.017 (95% CI 0.005, 0.062), p<0.001]. CONCLUSIONS/INTERPRETATION: Our data support the existence of a locus that modifies the protective effect associated with DQB1*0602. Typing for allele D6S265*15 can identify a less protective DQB1*0602 haplotype, thereby allowing a more accurate prediction of type 1 diabetes risk.

Progression to insulin-requiring diabetes in seronegative prediabetic subjects: the role of two HLA-DQ high-risk haplotypes.

AIMS/HYPOTHESIS: Most Caucasians with Type I (insulin-dependent) diabetes mellitus develop an autoimmune form of diabetes known as Type IA diabetes, based on the presence of humoral responses to islet autoantigens. Alleles at the HLA locus account for the strongest susceptibility to this form of diabetes, which requires insulin therapy. Because a number of patients who develop insulin-requiring diabetes are islet autoantibody negative, the HLA class II haplotypes, DQA1*0501-DQB1*0201 and DQA1*0301-DQB1*0302, were evaluated to assess whether they are an independent risk factor for progression to insulin requirement in first-degree relatives of Type I diabetic patients. METHODS: Both HLA-DQ genotyping and islet cell autoantibody assessment (insulin, GAD65, IA-2 autoantibodies and cytoplasmic islet cell antibodies) were evaluated prospectively in 74 relatives of Type I diabetic patients who developed diabetes treated with insulin (prediabetics) and in 426 control subjects who did not develop insulin-requiring diabetes. Based on the presence of DQA1*0501-DQB1*0201 and/or DQA1*0301-DQB1*0302, the number of HLA-DQ high-risk haplotypes was assigned as 0, 1 or 2. RESULTS: A higher prevalence of 2 HLA-DQ high-risk haplotypes was present in seronegative prediabetic subjects as compared to non-diabetic autoantibody negative first-degree relatives (33.3 % vs 10.1 % respectively; p < 0.05). Moreover, in seronegative relatives who developed insulin-requiring diabetes, the presence of 2 HLA-DQ high-risk haplotypes conferred an increased cumulative risk of developing insulin requirement of 27 % at 12.5 years of follow-up, compared to a risk of 6 % for non-diabetic relatives who were antibody-negative and had 0 or 1 HLA-DQ high-risk haplotypes (Log rank p = 0.01). CONCLUSION/INTERPRETATION: These data provide evidence for a phenotype, which is associated with the absence of conventional islet autoantibodies at initial screening, while usually remaining seronegative, and the presence of 2 HLA-DQ high-risk haplotypes with progression to clinical Type I diabetes after a prolonged follow-up. Given the fact that in humans the highest risk-conferring locus associated and linked to the disease is the HLA cluster, and that HLA-DQ molecules play a key role in the development of autoimmune diabetes, our observations imply that as yet unidentified immunologic abnormalities could well exist in seronegative relatives at risk of developing clinical diabetes and carrying 2 HLA-DQ high-risk haplotypes.

Menopause in type 1 diabetic women: is it premature?

Women with type 1 diabetes have a delayed menarche and a greater prevalence of menstrual disorders than women without diabetes. However, little is known about the menopause transition among type 1 diabetic women. The Familial Autoimmune and Diabetes (FAD) Study recruited both adult individuals who were identified from the Children's Hospital of Pittsburgh Type 1 Diabetes Registry for the years 1950-1964 and their family members. Unrelated nondiabetic control probands and their relatives were also evaluated. Women with type 1 diabetes (n = 143) compared with nondiabetic sisters (n = 186) or unrelated control subjects (n = 160) were more likely to have an older age at menarche (13.5, 12.5, and 12.6 years, respectively, P < 0.001), more menstrual irregularities before 30 years of age (45.7, 33.3, and 33.1%, respectively, P = 0.04), and a younger age at menopause (41.6, 49.9, and 48.0 years, respectively, P = 0.05). This resulted in a 6-year reduction in the number of reproductive years (30.0, 37.0, and 35.2 years, respectively, P = 0.05) for women with type 1 diabetes. Risk factors univariately associated with earlier menopause included type 1 diabetes (hazard ratio [HR] 1.99, P = 0.04), menstrual irregularities before 30 years of age (HR 1.87, P = 0.04), nulliparity (HR 2.14, P = 0.01), and unilateral oophorectomy (HR 6.51, P < 0.0001). Multivariate analysis confirmed that type 1 diabetes (HR 1.98, P = 0.056), menstrual irregularities by 30 years of age (HR 2.36, P = 0.01), and unilateral oophorectomy (HR 9.76, P < 0.0001) were independent determinants of earlier menopause in our cohort. We hypothesize that an earlier menopause, which resulted in a 17% decrease in reproductive years, is a major unstudied complication of type 1 diabetes.

Mortality trends in type 1 diabetes. The Allegheny County (Pennsylvania) Registry 1965-1999.

OBJECTIVES: To investigate long-term mortality and its temporal trends as of 1 January 1999 among the 1,075 patients with type 1 diabetes (onset age <18 years, diagnosed between 1965 and 1979) who comprise the Allegheny County population-based registry. RESEARCH DESIGN AND METHODS: Overall, sex- and race-specific mortality rates per person-year of follow-up were determined. Standardized mortality ratios were also calculated. Survival analyses and Cox proportional hazard model were also used. Temporal trends were examined by dividing the cohort into three groups by year of diagnosis (1965-1969, 1970-1974, and 1975-1979). RESULTS: Living status of 972 cases was ascertained as of January 1, 1999 (ascertainment rate 90.4%). The mean duration of diabetes was 25.2 +/- 5.8 (SD) years. Overall, 170 deaths were observed. The crude mortality rate was 627 per 100,000 person-years (95% CI 532-728) and standardized mortality ratio was 519 (440-602). Life-table analyses by the Kaplan-Meier method indicated cumulative survival rates of 98.0% at 10 years, 92.1% at 20 years, and 79.6% at 30 years duration of diabetes. There was a significant improvement in the survival rate between the cohort diagnosed during 1965-1969 and that diagnosed during 1975-1979 by the log-rank test (P = 0.03). Mortality was higher in African-Americans than in Caucasians, but there were no differences seen by sex. The improvement in recent years was seen in both ethnic groups and sexes. CONCLUSIONS: An improvement in long-term survival was observed in the more recently diagnosed cohort. This improvement is consistent with the introduction of HbA1 testing, home blood glucose monitoring, and improved blood pressure therapy in the 1980s.

High frequency of HLA-DQB1 non-Asp(57) alleles in Kuwaiti children with insulin-dependent diabetes mellitus.

The prevalence of polymorphic amino acids at position 57 of the HLA DQB1 in Kuwaiti children with insulin-dependent diabetes mellitus (IDDM) and nondiabetic controls has been determined using a polymerase chain reaction-sequence-specific primers (PCR-SSP) method. Using this approach, 34/55 (62%) IDDM children were found to be homozygous Ala/Ala and 19/55 (35%) were heterozygous with various combinations. Amongst the IDDM children with heterozygous genotype at codon 57 of HLA DQB1, 6/55 (11%) had Asp/Ala, 8/55 (15%) had Ala/Val, 4/55 (7%) had Ala/Ser and 1/55 had Asp/Val allelic combinations. When considered collectively, the nonaspartate (NA) alleles were represented in 87% of the IDDM cases and only 13% cases had Asp(57) allele in different heterozygous combinations, while none of the IDDM subjects had a homozygous Asp genotype. In nondiabetic controls, homozygous non-Asp (NA) alleles were represented in 44% subjects, 37% of the controls were heterozygous (NA/A) and 19% had a homozygous (A/A) genotype. These differences between the IDDM group and the control group were found to be statistically significant. Our data report one of the highest frequency of NA/NA residues at this locus compared with that from different world populations (Sardinians, Norwegians, US Caucasians, US Blacks and Chinese).

Type 1 diabetes in offspring of parents with type 1 diabetes: the tip of an autoimmune iceberg?

The objective of the present study was to examine the prevalence of self-reported autoimmune diseases among offspring of type 1 fathers, type 1 diabetic mothers, and non-diabetic parents. Type 1 diabetic probands (n=265; mean age=42 yr), who were ascertained from the Children's Hospital of Pittsburgh Registry for 1950-1964, recently participated in the Familial Autoimmune and Diabetes Study. Non-diabetic probands (n=96), identified from voter registration lists and matched by age, race, median income, and duration of residence in the Pittsburgh area, were also enrolled. Offspring of type 1 diabetic probands were more likely to have a reported autoimmune disease (5.8% vs. 2.4%; p=0.067) than offspring of non-diabetic probands. Half the cases in the diabetic families were disorders other than type 1 diabetes, (e.g., rheumatoid arthritis, Crohn's disease, etc.). Stratification by parental gender revealed a marginally higher risk for type 1 diabetes among offspring of type 1 diabetic fathers compared to mothers (4.9% vs. 3.4%; p=0.38, respectively, through age 20 yr). However, the risk for other autoimmune disorders was statistically significantly increased among offspring of type 1 diabetic mothers (0% vs. 6.2%; p=0.02, respectively, through age 20 yr). These data suggest that offspring of type 1 diabetic parents may be at high risk of developing other autoimmune disorders during childhood, with pediatric diabetes representing the 'tip of an autoimmune iceberg'. The observed risk differences by parental gender, which have also been reported for other autoimmune disorders, warrant further investigation.

Molecular epidemiology: the impact of molecular biology in epidemiology research.

Progress in molecular biology and genetics is changing the practice of medicine and public health through the development of molecular diagnostics and targeted interventions for susceptible individuals. The ethical, legal and social issues that are becoming apparent as these important discoveries are introduced into practice will have an enormous impact on society. The accurate translation of this new genetic information from the laboratory to the community is an urgent need. Molecular epidemiology is at the foundation of this important link, and represents the scientific basis of public health for the 21st Century.

A nine-year prospective study on the incidence of childhood type 1 diabetes mellitus in China.

To investigate the epidemiological characteristics of childhood type 1 diabetes mellitus in China, newly diagnosed cases of type 1 diabetes with an onset age under 15 years were retrospectively registered by 23 local centers in China following a standardized protocol on the basis of the nationwide registry established by the WHO DiaMond Project China Participating Center, Chinese Academy of Preventive Medicine (CAPM). A population of about 24 million children were covered in the defined areas. A two-sample capture-recapture method was used to estimate case ascertainment. Between 1988 and 1996, 903 diabetic cases were registered in 9 ethnic groups. The overall ascertainment corrected incidence rate (IR) was 0.59 per 100,000 person-year. The IR was 0.52/100,000 (95% CI: 0.50-0.54) for males and 0.66/100,000 (95% CI: 0.64-0.68) for females. The standardized ascertainment corrected IR by the national age-specific population in 1990 was 0.57 per 100,000 person-year. The incidence among various ethnic groups ranged from 0.25/100,000 to 3. 06/100,000. The IRs increased with northern latitude, and the IR of Han population was significantly higher in North China compared with South China (0.67 versus 0.53 per 100,000 respectively, P < 0.01). A correlation model of incidence and calendar time showed that the IR increased significantly between 1988 and 1996 (r = 0.86, P = 0.0027). The relative risk (RR) of type 1 diabetes mellitus for different age-groups estimated by a Poisson regression model showed that taking RR as 1.00 for age-group from 0 to 4 years, the RR for age-group from 5 to 9 year and from 10 to 14 year was 2.30 and 3.60 respectively. The standardized ascertainment corrected IR of childhood type 1 diabetes mellitus in China in much lower than in other countries. The geographic and ethnic variability of the incidence suggests that both genetic and environmental factors play a role in the development of childhood diabetes in China.

The HLA-DQ associations with Graves' disease in Chinese children.

OBJECTIVE: Graves' disease has been documented to be associated with different HLA genes in Caucasians and Chinese adults. The incidence of childhood Graves' disease has been reported to be high in Hong Kong Chinese. The aims of this study were to examine the HLA-DQA1 and DQB1 associations with Graves' disease in Chinese children. PATIENTS AND DESIGN: Sixty-seven Chinese children with Graves' disease (59 girls and 8 boys) and 51 racially matched healthy controls were recruited for the study. Genomic DNA was extracted from venous blood samples. HLA-DQ typings were determined by sequence-specific oligonucleotide probing of the respective enzymatically amplified gene. Frequencies of HLA-DQ alleles at each locus were compared between patients and controls using the chi 2-test. RESULTS: The frequency of HLA-DQB1.0303 was increased in the combined male and female patient group [53.7%; relative risk (RR) = 4.22; Pc = 0.005] and female patients (50.8%; RR = 3.76; Pc = 0.018) compared with that in the entire control group (21.2%). HLA-DQB1*201 was protective for Graves' disease (10.4%; RR = 0.20; Pc = 0.006). In contrast to studies in Caucasians, DQA1*0501 was not associated with susceptibility for Graves' disease in Chinese children. CONCLUSIONS: This study confirms that DQB1*0303 is a race-specific susceptible allele for Graves' disease in Chinese. Both susceptible and protective HLA-DQ alleles for Graves' disease in Chinese children are different from those in Caucasians.

Temporal trends in spontaneous abortion associated with Type 1 diabetes.

The objective of this study was to investigate temporal changes in the reported rates of spontaneous abortion associated with Type 1 diabetes. Individuals from the Children's Hospital of Pittsburgh Type 1 Diabetes Registry for 1950-1964 (n=495) completed a self-report reproductive history questionnaire in 1981 that was updated in 1990. Data from both surveys, which proved to be valid and reliable, were utilized for this report. More spontaneous abortions (26.8 vs. 7.7%, P<0.001), stillbirths (4.7 vs. 1.2%, P<0.001) and induced abortions (7.0 vs. 0.9%, P<0.001) were reported for Type 1 diabetic women than for the non-diabetic partners of Type 1 diabetic men. A significant temporal decline in the rates of spontaneous abortion for Type 1 diabetic women was observed (< or = 1969: 26.4%; 1970-1979: 31.0%; 1980-1989: 15.7%; P<0.05). No differences were apparent for the non-diabetic partners of Type 1 diabetic men (< or = 1969: 4.2%; 1970-1979: 9.5%; 1980-1989: 5.7%; P>0.05). Temporal changes in medical care for women with diabetes (i.e. self-monitoring of glycemic control) may have contributed to a recent reduction in spontaneous abortions associated with maternal Type 1 diabetes.

Familial resemblance of bone mineral density (BMD) and calcaneal ultrasound attenuation: the BMD in mothers and daughters study.

The familial resemblance in bone mineral density (BMD) and calcaneal broadband ultrasound attenuation (BUA) was examined in 207 mother-daughter pairs. Mothers were participants in the Study of Osteoporotic Fractures. Three groups of daughters were recruited based on their maternal history of "fracture," "low BMD" without fracture (< 0.58 g/cm2, t-score < -2.5), and "normal BMD" without fracture (> 0.67 g/cm2, t-score > -1.6). Data on other potentially heritable factors known to influence BMD and BUA were also collected. BMD was measured at the hip, spine, whole body, and calcaneus. Calcaneal BUA was assessed using the Walker-Sonix UBA 575. Total hip and femoral neck BMD were significantly lower (5.0-8.0%, p < 0.017) among daughters, in particular premenopausal daughters, of mothers with established osteoporosis ("fracture" or "low BMD") compared with daughters of mothers at lower risk of osteoporosis ("normal BMD"). BUA did not differ across daughter groups. Lifestyle characteristics (dietary calcium, smoking, physical activity) were not highly correlated in mothers and daughters. Height, weight, and body composition were significantly correlated within mother-daughter pairs and could be a potential mechanism by which BMD is inherited. Among pre- and postmenopausal daughters, heritability estimates ranged from 50-63% and 34-48%, respectively. Heritability for calcaneal BUA (53%) was evident among postmenopausal daughters only. In conclusion, familial association in BMD was strongest among premenopausal daughters. Estimates of heritability suggest that maternal BMD and BUA are important independent predictors of BMD and BUA among daughters, reinforcing the importance of prevention and early intervention among women with a positive family history of osteoporosis.

Islet cell autoimmunity in white and black children and adolescents with IDDM.

OBJECTIVE: To compare the frequency of islet cell antibodies (ICA) and antibodies to GAD65 and IA-2(ICA512) between black and white children and adolescents at the diagnosis of IDDM in a large consecutive series of cases from Children's Hospital of Pittsburgh. RESEARCH DESIGN AND METHODS: ICA and antibodies to GAD65 and IA-2 were measured in 437 white and black children and adolescents who were diagnosed with IDDM at < 19 years of age at Children's Hospital of Pittsburgh from January 1983 to December 1985, from January to December 1989, and from January 1996 to December 1997. RESULTS: The prevalence of ICA(H), GAD65, and IA-2 antibodies was significantly lower in blacks than whites at onset of the disease. In contrast, the prevalence of ICA(R) alone was higher in blacks. None of the antibodies were present in 12% of the blacks compared with 4% in whites. The same pattern was seen in both sexes. The prevalence of antibodies in white patients with onset of IDDM at <11 years of age was no different than in those who developed IDDM during adolescence. In contrast, black patients showed a significantly lower prevalence of almost all antibodies in the adolescent group. CONCLUSIONS: Black adolescents were more likely to not have antibodies, suggesting either that they have a nonautoimmune type of diabetes or that antibodies are not being detected by these assays.

Was there an epidemic of diabetes in nonwhite adolescents in Allegheny County, Pennsylvania?

OBJECTIVE: To determine the incidence of IDDM in children aged < 20 years at diagnosis in Allegheny County, Pennsylvania, for the period from 1 January 1990 to 31 December 1994 and to compare the incidence between whites and nonwhites in the same area and for the same time period. RESEARCH DESIGN AND METHODS: All new patients diagnosed between January 1990 and December 1994 who were aged < 20 years, on insulin, and residents of Allegheny County at diagnosis were identified from medical records of 23 hospitals in the Allegheny County area. To verify the completeness of the hospitals using the capture-recapture method, pediatricians and diabetologists were used as a secondary source. RESULTS: A total number of 257 patients were identified. The overall age-standardized incidence rate was 16.7/100,000. Nonwhites had a slightly higher incidence (17.6/100,000) than whites (16.5/100,000). In the 15-19 years age-group, the incidence in nonwhites (30.4/100,000) was almost three times higher than that in white (11.2/100,000) and more than two times higher than that in the previous period (from 1985 to 1989) (13.8/100,000). CONCLUSIONS: For the first time in the Allegheny County registry, and in any other registry, nonwhites showed a higher incidence of IDDM than whites. The high incidence in the 15-19 years age-group was responsible for this phenomenon. This epidemic of diabetes in adolescent nonwhites may be the result of a rising incidence of classical IDDM or another type of diabetes. Further studies using population-based registries are needed to determine whether this increase is being seen in other areas and other ethnic groups and to clarify the reasons for the increase in IDDM among blacks.

Hashimoto's thyroiditis and insulin-dependent diabetes mellitus: differences among individuals with and without abnormal thyroid function.

Insulin-dependent diabetes mellitus probands from the Familial Autoimmune and Diabetes Study were evaluated for autoimmune thyroid disease (n = 265). The prevalence of Hashimoto's thyroiditis was 26.6%; 42.0% of these individuals were euthyroid, and 58.0% were hypothyroid. There was a female predominance among hypothyroid and euthyroid Hashimoto's cases compared to those with no thyroid disease (75% vs. 72.4% vs. 41.6%; P < 0.001). Insulin-dependent diabetes mellitus patients with hypothyroid Hashimoto's thyroiditis were more likely to report another autoimmune disease compared to euthyroid Hashimoto's patients or individuals with no thyroid disease (30.8% vs. 17.2% vs. 13.9%; P < 0.01). Sex-specific analysis revealed that this difference was significant for men but not for women. Both euthyroid and hypothyroid Hashimoto's cases were more likely to have a family history of the disease (66.7% vs. 69.2% vs. 47.7%; P < 0.05). No differences were observed in the prevalence of DQA1*0501-DQB1*0201 or DQA1*0301-DQB1*0302 across the three groups. Body mass index, lipid levels, glycemic control, and diabetes complications were also similar. However, euthyroid Hashimoto's women were more likely to report spontaneous abortions than those with hypothyroid Hashimoto's thyroiditis or no thyroid disease (23.8% vs. 61.5% vs. 29.1%; P < 0.05). These data suggest that gender-specific risk factors may be primary determinants of Hashimoto's thyroiditis and other autoimmune diseases among women. However, disease-specific determinants may also increase susceptibility to other autoimmune diseases.

Childhood diabetes in China. Enormous variation by place and ethnic group.

OBJECTIVE: To investigate the incidence rate of IDDM in China. RESEARCH DESIGN AND METHODS: The Chinese IDDM registry was established in 1991 as part of the World Health Organization's Multinational Project for Childhood Diabetes (DiaMond) project. Twenty-two centers were developed to monitor the incidence of IDDM in children < 15 years of age. The population under investigation includes > 20 million individuals, representing approximately 7% of the children in China. Capture-recapture methods were used to estimate the ascertainment. RESULTS: The overall ascertainment-corrected IDDM incidence rate in China was 0.51 per 100,000, the lowest rate ever reported. There was a 12-fold geographic variation (0.13-1.61 per 100,000). In general, the incidence rate was higher in the north and the east. There was a sixfold difference among ethnic groups (highest: Mongol group, 1.82 per 100,000; lowest: Zhuang group, 0.32 per 100,000). CONCLUSIONS: China has an extremely low overall IDDM incidence rate. China also has the greatest geographic and ethnic variation seen for any country.

IDDM incidence in a multiracial population. The Hawaii IDDM Registry, 1980-1990.

OBJECTIVE: The Hawaii IDDM Registry was created to determine the incidence rate of IDDM among children aged < 15 years of Oahu between 1980 and 1990. Because of the multiracial population living in Hawaii, it is an ideal state in which to study the effect of migration on IDDM incidence. RESEARCH DESIGN AND METHODS: Data were collected by a retrospective hospital chart review and physician survey. Registry criteria included 0-14 years of age at IDDM diagnosis and primary residence on Oahu. Children who were military dependents were excluded. Denominator data were compared from two sources. RESULTS: A total of 113 new-onset IDDM cases were identified. Case ascertainment was 97%. The unadjusted annual incidence rate was 7.04-7.61 per 100,000 (95% CI 5.83-9.19), depending on which denominator source was used. Race- and ethnicity-specific rates varied greatly (all rates per 100,000): part Hawaiian, 15.34-16.58; Caucasian, 6.21-6.71; Filipino, 3.66-3.96; and Japanese, 2.85-3.08. Temporally, the incidence fluctuated between a low of 2.96 per 100,000 in 1981 to highs of 11.11 and 11.85 per 100,000 in 1985 and 1989, respectively. Ascertainment-corrected rates for these years (3.70, 11.76, and 13.48 per 100,000, respectively) show that the fourfold incidence increase between 1980 and 1989 was not due to ascertainment differences. CONCLUSIONS: IDDM incidence among children < 15 years of age in Hawaii was the lowest documented in the U.S. The incidence among part-Hawaiian children was 2.5 times greater than that of Caucasian children in Hawaii. IDDM incidence rates among Japanese children in Hawaii were comparable with rates in Japan. The temporal variation in IDDM incidence corresponded with a mid-1980s pandemic of IDDM documented elsewhere.