Seasonal variation in the diagnosis of type 1 diabetes

Abstract

In response to the report of Samuelsson et al. [ [1] Samuelsson U. Carstensen J. Löfman O. Nordfeldt S. Seasonal variation in the diagnosis of type 1 diabetes in south-east Sweden. Diab. Res. Clin. Pract. 2007; 76: 75-81 Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar ] we like to add to the seasonal variation in the diagnosis of type 1 diabetes from our data between 1992 and 2001 of the Dutch Pediatric Surveillance Unit (1284 newly diagnosed children of 0–14 years old in this period). The seasonality of first insulin injection and date of birth was studied as well as space–time clustering by Mantel's statistic (Z) for first insulin injection and date of birth. We also found the number of first insulin injections to be highest during winter months, and lowest during summer time (χ2 = 52.71, p < 0.001). Date of birth of newly diagnosed patients vary, in comparison with the distribution of all births, differences are not significant. Mantel's statistic (Z) for space–time clustering shows significant clustering at first insulin injection, and clustering at birth (respectively, p = 0.01 and <0.01). Our results are consistent with the viral hypothesis. Thus it could be that contact with infectious agents at birth as well as later in childhood, especially during the winter season, occurs before the onset of the manifestation of the disease [ [2] Van Wouwe J.P. Jacobusse G. Reeser H.M. Van Buuren S. Seasonality and time–space clustering for date of birth and onset of childhood type 1 diabetes (T1D) in the Netherlands between 1992 and 2001. Eur. J. Clin. Invest. 2004; 34: 7 Google Scholar ].