Incidence of type I diabetes in the Liguria Region, Italy. Results of a prospective study in a 0- to 14-year age-group

R T R E P O R T Incidence of Type I Diabetes in the Liguria Region, Italy Results of a prospective study in a 0- to 14-year age-group MASSIMO MAZZELLA, MD MARIO COTELLESSA, MD STEFANO BONASSI, PHD ROBERTA MULAS, MD ANTONINO CARATOZZOLO, MD SHAMIR GABER, MD CESARE ROMANO, MD OBJECTIVE — To assess updated incidence of insulin-dependent diabetes mellitus (IDDM) in 0- to 14-year-old children in Liguria, a northwest region of Italy. RESEARCH DESIGN A N D M E T H O D S — Incident cases were recorded prospectively from 1987 to 1991. Incidence rates (IRs) were directly standardized on the basis of the 1990 world population. The independent effect of age, sex, residence, and calendar year was estimated with a Poisson regression model. The degree of ascertainment was calculated in accordance with the capture/recapture method. RESULTS — During 5 full calendar years, 117 new cases of IDDM in children were diagnosed in Liguria. The standardized IR over the 5-year period was 11.72 cases • 100,000' l • year"1. The sex-specific IR among males and females was 11.45 and 12.01, respectively. The age-specific IR was higher in the 5-9 age-group. CONCLUSIONS — The IR of IDDM in Liguria is among the highest in southern Lurope and approaches that of northern European countries. In particular, it is much higher than those reported in the surrounding Italian regions, except for Sardinia. Therefore, the geographical distribution of IDDM does not seem to reflect the simple north-south gradient reported in several previous studies. T he fascinating puzzle and unsolved questions about the etiology and pathogenesis of insulin-dependent diabetes mellitus (IDDM) have recently emphasized the importance of epidemiological studies. This research may contribute to clarifying the role of both predisposing, From the University Department of Pediatrics, G. Gaslini Scientific Institute and Department of Environmental Epidemiology, Istituto Nazionale per la Ricerca sul Cancro, Genoa, Italy. Address correspondence and reprint requests to Mario Cotellessa, MD, Regional Reference Pediatric Center for Diabetology, University Department of Pediatrics, G. Gaslini Scientific Institute, Largo Gerolamo Gaslini 5, 16148 Genoa, Italy. Received for publication 23 August 1993 and accepted in revised form 12 May 1994. IDDM, insulin-dependent diabetes mellitus; IR, incidence rate; RAN, Regional Assistance Network; CI, confidence interval; RR, relative risk. DIABKTES CARE, VOLUME 17, NUMBER 10, OCTOBER 1994 initiating, and/or precipitating factors of IDDM, all of which may have a role in the development of the disease. The main purpose of this study is to provide current and reliable data about the incidence rate (IR) in the Liguria region, as well as to evaluate the role of some external factors such as sex, age, calendar year, or residence in explaining the occurrence of IDDM. RESEARCH DESIGN A N D M E T H O D S — IDDM patients who were diagnosed between 1 January 1987 and 31 December 1991 and who were <15 years of age at the time of their (irst injection were included in the study. Diagnosis of IDDM was in accordance with the National Diabetes Data Group criteria (1). It was assumed that two alternative sources provided the information independently. Data from a primary source were obtained from records of 9 pediatrics departments, 11 endocrinology and diabetes units, and 8 internal medicine departments in the hospitals of the four provinces of the region (Genoa, La Spezia, Savona, and Im • peria), assuming that all patients were hospitalized at the onset of insulin therapy. Actually, the possibility that a child with onset of IDDM <15 years of age could have been treated without any contact with a hospital department must be regarded as exceptional in Liguria and in Italy. We sent a study questionnaire and a letter to the chiefs of the departments explaining the aim of the study and the data required (name, age, sex, residence, date of first insulin injection, parents' ages and professions, family history for IDDM, non-insulin-dependent diabetes mellitus, or other endocrinopathies, symptoms at onset, and intercurrent infection or stress in the last 3 months before diagnosis.) The research was made easier because of the Regional Assistance Network (RAN) for IDDM. This collaborative network was set up in 1986 under the supervision of the Pediatric Depart- 1193 Type I diabetes in the Liguria region Table 1—IR (per 100,000/year) oflDDM in Liguria during 1987-1991 Males Age-group (years) 0-14 0-4 5-9 10-14 Females Age-group (years) 0-14 0-4 5-9 10-14 Both sexes Age-group (years) 0-14 0-4 5-9 10-14 Population Cases Rate 95% Cl 499,500 141,933 152,000 205,767 59 12 20 26 11.45* 8.46 13.16 13.12 8.6-14.5 3.7-13.2 7.4-18.9 8.2-18.1 472,800 132,650 144,663 195,487 58 11 23 25 12.01* 8.29 15.90 12.28 9.0-15.3 3.4-13.2 9.4-22.4 7.4-17.2 972,500 275,582 296,663 401,254 117 23 43 51 11.72* 8.37* 14.38* 12.91* 7.6-16.2 6.1-10.6 11.7-17.4 10.4-15.0 *Standardized rates ment of the University of Genoa to provide current and reliable epidemiological data of IDDM in childhood in the different provinces of Liguria by an active surveillance model. The secondary source of validation was established through the review of IDDM patients' basic data (name, surname, date, and place of birth) in the 20 local health service units in the region. Our center was notified about each new case, and each case was personally identified to avoid multiple reports of the same patient. Age- and sex-specific IRs were calculated for the four Ligurian provinces. Census data are collected in Italy the first year of each decade, so in our study, we used real census data only for 1991. The estimates of the 1987-1990 resident population were provided directly by the Italr ian Institute of Statistics. IRs were directly standardized on the basis of world population (2). The independent effect of age, sex, residence, and calendar year was estimated with a Poisson regression model (3). The choice among different regres- 1194 sion models was made as suggested by Kleinbaum et al. (4). The calculation of rate ratios and 95% confidence intervals (CIs) was performed by the Generalized Linear Interactive Modeling statistical software (5). Completeness of ascertainment was estimated with the capture/ recapture method (6). RESULTS— During 5 calendar years, 117 new cases of IDDM who met all eligibility criteria were identified. Out of these, 15 (12.8%) were found by primary source only, nine (7.6%) by secondary source only, and 93 (79.4%) by both sources. Combination of both sources provided a degree of ascertainment of 98.77%. The age- and sex-standardized IR in 1987-1991 was 11.72 cases • 100,000" l • year"1. The theoretical IR (crude IR/degree of ascertainment) was 12.17 cases • 100,000"* • year"1. Number of eligible cases, population, and standardized IRs by sex and age-groups from 1987 to 1991 are reported in Table 1. No significant differences were observed according to age or sex in the crude or standardized rates (data not shown). All models included sex and age as potential confounding factors. No covariate was shown to affect significantly the IR of IDDM. Among single categories, only age class 5-9 showed a significant excess (relative risk [RR] = 1.70, 95% CI 1.02-2.82) when compared with the 0-4 age-group (Table 2). IRs by year of diagnosis were examined, but no significant variation was seen during 5 years. The highest IR was reported in 1988 (14.69), and the lowest was in 1990 (10.48). Genoa and Savona provinces showed the highest rates (12.30 and 12.35, respectively), whereas La Spezia and Imperia showed the lowest rates (9.36 and 8.97, respectively). The variation in incidence among the four provinces of the region was not statistically significant. The highest rates were found in the 5-9 and 10-14 age-groups (14.38 and 12.91, respectively), and the lowest rate was found in the 0-4 agegroup (8.37). Regarding the sex differences in age subgroups, female rates were higher than male rates in the 5—9 agegroup and lower in the 10-14 and 0-4 age-group. No statistical differences were observed. Overall incidence was somewhat higher for females than for males (12.01 vs. 11.45), but this finding failed to reach statistical significance. The number of incident cases gradually increases during childhood (23, 43, and 51 cases in 0-4, 5-9, and 10-14 age-groups, respectively). The percentage of distribution of cases by month of diagnosis showed a winter time peak. The highest number of new cases was recorded in March, and the lowest number was recorded in June. There was no significant difference in seasonal pattern. CONCLUSIONS— Monitoring the occurrence of IDDM is critical for a better understanding of the disease. Only recently, within the framework of the DIABETES CARH, VOII:MH 17, NI:MBI-:R 10, Oc TOBIR 1994 Mazzella and Associates Table 2—Results of fitting Poisson regression model to data on diabetes incidence in Liguria Covariates Sex F M Age (years) 0-4 5-9 10-14 Province Genova Imperia Savona La Spezia Calendar year 1987 1988 1989 1990 1991 Log likelihood ratio test Degree of freedom P 0.03 1 0.858 2 4.47 2.75 DIABETES CARE, VOLUME 17, NUMBER 10, OCTOBER 1.00 1.03 0.72-1.48 1.00 1.70 1.44 1.02-2.82 0.88-2.36 1.00 0.66 1.00 0.73 0.35-1.24 0.61-1.62 0.41-1.32 1.00 1.33 1.11 0.97 1.06 0.77-2.30 0.62-1.98 0.53-1.76 0.59-1.90 — — 0.432 — 0.811 4 project EURODIAB (7), has standard protocol for prospective surveillance of the incidence of IDDM been available, but we are a long way from reliable information. In Italy, in particular, previous epidemiological studies have been conducted (8,9), but the results are not comparable because of the heterogeneous methodology, the difficulties in coordinating different centers, and the lack of a central register. In Liguria, the RAN for IDDM has partially overcome these difficulties. RAN is a collaborative network set up in 1986 under the supervision of the Pediatric Department of Genoa University and performed through a strict cooperation with pediatric diabetologists in the region to 1) provide current and reliable epidemiological data of IDDM in childhood in the different provinces of Liguria by an active surveillance model and 2) make uniform and update diabetes-care standards and guidelines. The RAN represents the first experience in Italy of a collaborative net- 95% Cl 0.106 3 1.58 RR — work for IDDM using a validated epidemiological methodology. The degree of ascertainment is very high (98.7%), allowing a reliable assessment of IDDM incidence for subjects <15 years of age in Liguria. A comparison with available rates in other countries shows that our region has an intermediate IDDM risk. IR in Liguria is lower than those reported in Finland, Sweden, Norway, Denmark, and Scotland (7), but it is similar to those of other northern European countries such as the Netherlands, Belgium, Luxembourg (7), Iceland (10), and the British Isles (11), and it is among the highest in southern Europe excluding Sardinia (12). These data offer an exception to the hypothesis that latitude (north-south gradient) may be critical to explain the considerable geographical variation of IDDM within Europe and throughout the world (7,13). Moreover, it is a surprising finding that, despite the supposed genetic homogeneity and very similar environmental conditions, diet, and lifestyle, Liguria shows an IR approximately dou- 1994 ble those reported in other Italian regions such as Lombardia (7), Lazio (7), and Campania (14). This finding represents a challenging clue for further epidemiological research to provide more information about the distribution of genetic markers and the role of environmental factors in etiology. The IR remained fairly stable during 5 years of the study. This is not consistent with some reports that suggest a short-term increase in IDDM IRs (15-17). Nevertheless, there is evidence of a longer-term temporal increase in IDDM incidence in northern Europe (equivalent to a doubling over the past 20 years) (18). Because our study encompasses only 5 years, the examination of temporal trend may provide little information, so we are unable to say whether our rate reflects a steady pattern or the evidence of the latest figures of a long-term rising incidence. Rates over the next 5-10 years will be more informative, so further research is warranted. In accordance with other authors, we found that the onset of IDDM is affected by seasonality, although the variability is not large (19). A higher incidence of viral infections (20) during the winter season or climatic factors, such as cold, which might influence the time of onset of IDDM by increasing the need for insulin, supports this hypothesis. Regarding the age of onset, in both males and females, the peak age at diagnosis is around puberty, between 11 and 12 years (earlier in females) in agreement with the general experience (7,12). This is consistent with the typical pattern of occurrence of IDDM by age that shows a striking association between the peak of incidence and the pubertal growth spurt (21). In conclusion, this prospective 5-year study has shown an IR of IDDM in Liguria as among the highest in southern Europe, approaching those reported in several northern European countries. On this basis, Liguria represents a new exception to north-south gradient and may contribute to drawing a more complete 1195 Type I diabetes in the Liguria region ACE study. Lancet 339:905-909, 1992 8. Cerutti F, Bruno A, Bruno E, Casetta F, Balboni R, Urbino A: Incidenza e prevalenza del Diabete mellito insulino-dipendente ad esordio infanto-giovanile nella Acknowledgments— This work was supported by Health Councillorship of the Liguria citta di Torino e provincia. Minerva PediRegion. atr 37:67-70, 1985 9. DERI Study Group: The epidemiology and immunogenetics of IDDM in Italian References heritage population. Diabetes Metab Rev 1. National Diabetes Data Group: Classifica6:63-69, 1990 tion and diagnosis of diabetes mellitus 10. Thorsson AV, Helgason T, Jonasson MR, and other categories of glucose intolerDanielsen R: Childhood diabetes in Iceance. Diabetes 28:1039-1049, 1979 land in the last two decades (Presenta2. United Nations: World Population Prospects tion). International Symposium on Epi1990. New York, United Nations, 1991 demiology and Etiology of IDDM in the 3. From EL, Checkoway H: Use of Poisson Young, France, May 1991 regression models in estimating incidence 11. Metcalfe MA, BaumJD: Incidence of insurates and ratios. Am] Epidemiol 121:309— lin-dependent diabetes in children under 323,1985 15 yrs in the British isles during 1988. Br 4. Kleinbaum DG, Kupper LL, Morgenstern Med] 302:443-447, 1991 H: Epidemiology Research. New York, Van 12. Muntoni S, Songini M, Sardinian CollabNostrand Reinhold, 1982 orative Group for Epidemiology of IDDM: 5. Baker NJ, Nelder JA: Generalized Linear InHigh incidence rate of IDDM in Sardinia. teractive Modeling (GLIM), Release 3. Oxford, Diabetes Care 15:1317-1322, 1992 U.K., Numerical Algorithms Group, 1978 13. Diabetes Epidemioloy Research Interna6. Hamman RF, Gay EC, Cruickshanks KJ, tional Group: Geographic patterns of Cook M, Lezotte DC, Klingensmith GJ, childhood insulin-dependent diabetes Chase HP: Colorado IDDM registry: incimellitus. Diabetes 37:1113-1119, 1988 dence and validation of IDDM in childhood aged 0-17 yrs. Diabetes Care 13: 14. Prisco F, Iafusco D, Palumbo F, Sagliocca L, Vicedomini D, Boccia C, Amodeo B, De 499-506, 1990 Felice D, Stoppoloni G: Incidenza in 7. Green A, Gale EAM, Patterson CC: InciCampania del Diabete Mellito Tipo 1, in dence of childhood onset insulin-depensoggetti di eta inferiore a 14 aa nel triendent diabetes mellitus in the EURODIAB map of geographical distribution of childhood-onset diabetes in Europe. 1196 15. 16. 17. 18. 19. 20. 21. nio 1989-1991. II Diabete (Suppl. 2): 417-418, 1991 Huff JC, Hierholzer JC, Farris WA: An outbreak of jzuvenile diabetes mellitus: consideration of a viral etiology. Am] Epidemiol 100:277-287, 1974 Rewers M, La Porte RE, Walczak M, Dmochowski K, Bogazcynska E: Apparent epidemic of insulin-dependent diabetes mellitus in midwestern Poland. Diabetes 36:106-113, 1987 World Health Organization DIAMOND Project Group on Epidemics: Childhood diabetes, epidemics and epidemiology: an approach for controlling diabetes. Am J Epidemiol 135:803-816, 1992 Bingley PJ, Gale EAM: Rising incidence of IDDM in Europe. Diabetes Care 12:289295,1989 Gamble DR, Taylor KW: Seasonal incidence of diabetes mellitus. Br Med J 3:631-633, 1969 TuomilehtoJ, Lounamaa R, TuomilehtoWolf E, Reunanen A, Virtala E, Kaprio EA, Akerblom HR, the Childhood Diabetes in Finland Study Group: Epidemiology of childhood diabetes mellitus in Finland: background of a nationwide study of type 1 (insulin-dependent) diabetes mellitus. Diabetologia 35:70-76, 1992 Dahlquist G, Blom L, Holmgren G, Hagglog B, Jarsson Y: The epidemiology of diabetes in Swedish children 0-14 years: six years prospective study. Diabetologia 281:802-808, 1985 DIABETES CARE, VOLUME 17, NUMBKR 10, OC.IOBIR 1994