The incidence of childhood type 1 diabetes showed a worldwide increase towards the end of the 20th century. Earlier sources of information are less robust, but suggest that childhood diabetes was rare prior to the introduction of insulin, and had a low incidence in all parts of the world before the mid-century. Several countries then observed a parallel increase in the 1950s and 1960s, with some suggestion of a delayed upswing in other populations. Although local patterns vary to some extent, the overall pattern has been one of linear increase with occasional plateaus. No epidemics (a sharp rise in incidence followed by a fall) have been convincingly demonstrated. Both sexes are equally affected before puberty in high incidence countries, but there is a modest male excess thereafter. The most rapid percentage increase from baseline has occurred in the under-5 age group, although the rate of increase is similar in children aged 5–9 and 10–14 years. Genetic comparisons between long-term survivors of type 1 diabetes and recently diagnosed children suggests that the increase has largely occurred in those with markers of HLA susceptibility, but that these have shown increased penetrance in the younger age groups.
The changing demography of childhood diabetes has major implications for our understanding of the disease. A rapid change in incidence within a genetically stable population implies that non-genetic factors are active and that the influence of genes is relative to population, time and place. It suggests that something has changed in the environment our children encounter or in the way they are reared. Understanding this historical change would open the way to rational forms of intervention, which could be introduced at the stage of development when they are most likely to prove effective. Seen from this perspective, the central task of diabetes prevention is to understand a historical trend, and to put it in reverse.
In 1913, the Professor of Pediatrics at the Harvard Medical School had personal knowledge of 19 cases of childhood diabetes, and found a total of 989 cases in the world literature since 1878. At that time 80–90% of children diagnosed with diabetes died within 2 years, mostly in diabetic ketoacidosis. Since childhood mortality was high in the background population, the condition was rare, and diagnostic facilities limited. Many others must have died from undiagnosed diabetes.
A longitudinal survey in Oslo from 1925 to 1954 suggested that the incidence of diabetes under the age of 30 years was stable and low at about 1/100,000 per year; of interest, the incidence of diabetes dropped in older age groups as a result of wartime rationing, but the incidence in the young was unaffected.
The second figure shows the incidence of childhood diabetes in Norway through to the end of the century. This suggests an increase in incidence from the mid-century, with the suggestion of a plateau over its last decade. Data from Denmark, Finland, Sardinia, the UK and the USA would also be consistent with a mid-century rise, although better access to diagnostic facilities may have made some contribution to the apparent increase. Countries with a current high incidence may thus have been the first to witness a rising incidence.
In support of this, recent data show that incidence is rising most rapidly in countries in central and eastern Europe, which had low incidence rates in the past, and are tending to converge on the higher rates seen in western Europe.
Current patterns of increase
The rising incidence of type 1 diabetes came to general notice in the 1980s, and has been abundantly documented in all parts of the world since then. In Europe, the overall annual incidence rose by an average of 3.9% annually from 1989 to 2003, with a log-linear pattern of increase in most environments.
Although the rate of increase runs in parallel in the 0–4, 5–9 and 10–14 year old age groups, the most rapid rise from baseline is seen in the under 5s, and is predicted to translate into a doubling in the numbers diagnosed under the age of 5 years between 2005 and 2020, and an overall increase of 60% in the prevalence of childhood diabetes over the same period.5
Finland has the highest incidence of childhood type 1 diabetes in the world, followed by Sardinia. Recent data showed an increase in the age-standardized increase from 31.4/100,000 per year in 1980 to 64.2/100,000 per year in 2005. The number of new cases under the age of 15 is expected to double in 15 years, and the rate of increase currently appears to be accelerating.
Why the increase?
The increasing incidence of childhood type 1 diabetes cannot plausibly be explained by transmission distortion (increased transmission of a susceptibility gene or genes from one generation to the next). Susceptibility genes passed on by parents with type 1 diabetes to their children can account for only a small fraction of the observed increase. The risk of type 1 diabetes in the offspring rises slightly with increasing maternal age at delivery, and the trend for women to start their families later may also be responsible for a small fraction of the increase.
It should not be forgotten that incidence rates for type 1 diabetes are largely based upon observations in children under the age of 15 years. A few registries extend ascertainment to 30 or 40 years of age, but the true lifetime incidence of type 1 diabetes is unknown. This gives rise to two possibilities: (1) That there has been a true increase in the proportion of the population developing type 1 diabetes; or (2) that the proportion developing type 1 diabetes is unchanged, but the condition now presents earlier.
The second of these has been called the 'spring harvest' hypothesis. Support for this interpretation comes from the genetic studies, to be discussed below, and from studies in Sweden and Belgium in which the increase in diagnosis under the age of 15 years has been cancelled by a decrease in the number diagnosed over that age. The reason for this increase is unexplained, although the trend towards earlier puberty might be expected to result in an earlier peak incidence of diabetes, and thus account for another small fraction of the observed change.
Evidence for the changing penetrance of susceptibility genes
Diseases with a strong genetic component tend to present earlier in life, and type 1 diabetes is no exception. Thus, heterozygotes carrying the two HLA susceptibility alleles HLA-DR3/DR4 are over-represented in the under-5-year-old age group, and are less frequently encountered with increasing age. Comparisons between long-term survivors of type 1 diabetes and recently diagnosed children, matched for age of onset, show that the proportion with the highest risk susceptibility genotype in each childhood age band has diminished over time, suggesting greater penetrance of diabetes susceptibility genes, most likely in response to an environment that is more permissive for the onset of diabetes. 
Genetic susceptibility could thus be pictured as a rock surrounded by water, representing environmental protection against the development of diabetes. When protection is high, only the highest risk genes will be expressed; conversely, reduced protection will lead to earlier expression of the disease in genetically susceptible individuals.
Possible explanations for decreased environmental protection against the development of diabetes include the concept that overnutrition and increasing insulin resistance promote autoimmunity and earlier onset of diabetes (the Accelerator hypothesis), and the concept that decreased exposure to antigenic stimuli such as infections and parasites in early life results in delayed maturation of the immune defensive system and increased susceptibility to both allergic and autoimmune disease (the hygiene hypothesis). These are reviewed elsewhere in this section.
^ Gale EAM. The rise of childhood diabetes in the 20th century. Diabetes 2002;51:3353–61
^ Morse JL. Diabetes in infancy and childhood. Boston Medical and Surgical Journal 1913;168:530–35
^ Westlund K. Incidence of diabetes mellitus in Oslo, Norway, 1925 to 1954. Br J Prev Soc Med 1966;20:105–16
^ Krolewski AS et al. Epidemiologic approach to the etiology of type 1 diabetes and its complications. N Engl J Med 1987;317:1390–8
^ Patterson CC et al. Incidence trends for childhood type 1 diabetes in Europe during 1989-2003 and predicted new cases 2005-2020 - a multicentre prospective registration study. Lancet 2009;373:2027–33
^ Harjutsalo V et al. Time trends in the incidence of type 1 diabetes in Finnish children - a cohort study. Lancet 2008;371:1777–82
^ Pitkaniemi J et al. Increasing incidence of type 1 diabetes - role for genes? BMC Genetics 2004;5:5
^ Gale EAM. Spring harvest? Reflections on the rise of type 1 diabetes. Diabetologia 2005;48:2445–50
^ Hermann R et al. Temporal changes in the frequencies of HLA genotypes in patients with type I diabetes—indication of an increased environmental pressure? Diabetologia 2003;46:420–5
^ Gillespie KM et al. The rising incidence of type 1 diabetes is associated with a reduced contribution from high-risk HLA haplotypes. Lancet 2004;364:1699–1700