Incidencija, klinička prezentacija i regionalne razlike u tipu 1 šećerne bolesti u djece u dobi od 0 do 14 godina u Hrvatskoj: devetogodišnje praćenje

Aims/Hypothesis: The aim of this study was to determine incidence and trends of Type 1 diabetes in children aged 0-14 years in Croatia from 1995 to 2003. Further, to determine differences in incidence trends according to age groups 0-5, 5-9 and 10-14 years, and to evaluate regional differences in incidence and trends of incidence as well as modes of clinical presentation with special attention to diabetic ketoacidosis (DKA), and possible changes in means of clinical presentation during the nine years of follow up. Methods: The first source was data on newly diagnosed patients from pediatric departments from the country, and the second source were records of children with Type 1 diabetes from Records of the Croatian Diabetes Association. The ascertainment was estimated with capture-recapture method. Population data were obtained from the Croatian Bureau of Statistics based on the population censuses in 1991 and 2001, and based on the annual estimates for periods between the censuses. The crude incidence rates were expressed as the number of newly diagnosed Type 1 diabetes patients per 100 000 person-years for three age groups (0-4, 5-9, and 10-14 years) and for the group as a whole (0-14 years) separately for boys and girls. The method of direct standardization to standard world population for age 0-14 years was used for calculation of standardized incidence. Poisson regression model was used to analyze changes in incidence and trends of Type 1 diabetes mellitus for period from 1995 to 2003 and to examine the incidence rate differences among the age groups, gender differences, and for potential regional differences. Method of logistic regression and Cochran-Mentel-Haenszel test were used to analyze factors with potential influence on DKA incidence. Trend of DKA incidence in the nine year period was assessed using Cochran-Armitage trend test. Characteristics from the onset of the disease were evaluated using One Way Analysis of Variance (ANOVA). Results: 692 children were diagnosed with Type 1 diabetes mellitus, out of which 323 were girls and 369 were boys. Using capture-recapture method completeness of ascertainment was estimated to 97,4%. Standardized incidence for the whole age group is 8,87/100000/year (CI 95% 5,07-12,68), for girls 8,47 (CI 95% 7,54-9,41), for boys 9,26 (CI 95% 8,30-10,21). Although the incidence was higher in boys than in girls, the effect of gender on the incidence was not statistically significant. For the whole age group during the evaluated period there was an upward trend of incidence (χ2 32,6; P <0,001) Type 1 diabetes. Annual increase in incidence was 9% (95% CI: 5,8%-12,2%). According to age, significance of upward trend of incidence was shown in all age groups [0-4 years 14,0% (95% CI: 8,8%-22,52%), (χ2 14,89; P 0,0001); 5-9 years 8,3% (95% CI: 3,2%-13,7%), (χ2 10,36; P 0,00013); 10-14 years 7,7% (95% CI: 3,1%-12,7%), (χ2 10,76; P 0,00010)]. There was no significant difference in incidence trends among age groups (χ2 2,26; P 0,32). Divided by regions, the highest incidence was recorded in Southern Croatia 10,91 (95% CI 9,44-12,6) per 100000/year while lower incidence was recorded in Eastern Croatia 8,93 (95% CI 7,44-10,71) per 100000/year, and in Central Croatia 8,64 (95% CI 7,59-9,84) per 100000/year. Significant difference in incidence was reported between Central and Southern Croatia (χ2 5,63; P 0,017). All three regions note upward trend of incidence, but the trend significance is present in Eastern [11,4% (95% CI 3,9-19,4), (χ2 9,42; P 0,0021)] and Southern [7,2% (95% CI 1,8-12,9), (χ2 7,03; P <0,008)], but not in Central Croatia [4,6% (95% CI 0,0-9,8), (χ2 3,3; P 0,06)]. According to age in all three regions the upward trend was statistically significant in the youngest age group. The significance of trend in older age groups was persistent only in Eastern Croatia. Duration of symptoms prior to diagnosis of Type 1 diabetes is less than 3 weeks. Polyuria, polydipsia and weight loss are the leading symptoms. In effort of early disease recognition more emphasis should be given to non-classical symptoms such as secondary enuresis which frequently occurs in children younger than 10 years. The incidence of DKA at the disease onset was 36,4% regardless of age and gender, but regional differences were noted with highest incidence of DKA in Central Croatia (41,7%) which is significantly larger than in Southern Croatia (31,28%), (χ2 4,41; P 0,0357 ). Number of children who at diagnosis of Type 1 diabetes have DKA is decreasing during the nine year period from 41,67% to 33,33%, with statistical significance (Z 1,68; P 0,046). Negative family history, meaning disease free first line relatives, increases by 2,4 times the chance of DKA being the first presentation of Type 1 diabetes (χ2 5,89; P 0,015). Conclusion: The incidence rate places Croatia among a group of countries with intermediate risk for development of Type 1 diabetes mellitus. In the past nine years an upward trend of incidence was noted. The significance of this increase is present in all age groups, especially in the youngest group up to five years of age. Further studies will reveal whether this upward trend will continue. Among regions there were differences in incidence, as well as in statistical significance of incidence trend. The highest incidence was recorded in Southern Croatia. All three regions show upward trend of incidence, but the significance of trend is present in Eastern and Southern, but not in Central Croatia. According to age in all three regions the upward trend was statistically significant in the youngest group. The significance of trend in older age groups was conserved only in Eastern Croatia. Further follow up is needed to determine whether the regional differences are consequence of population dynamics from the previous period or whether they will continue and possibly implicate possible population genetic differences or environmental factors influencing the development of the disease. Number of children who present with DKA at disease diagnosis decreases during the nine year period, but is still high. This is why in every, especially in younger, acutely ill child at diagnosis of unclear disease one should evaluate the possibility of Type 1 diabetes to avoid DKA. More intensive education on early signs of the disease would help general population as well primary care health workers.