Diabetes mellitus (DM) is a major global public health issue as it imposes a heavy burden on mortality and disability, as well as on socio-economic development [1]. DM is one of five priority non-communicable diseases (NCDs) targeted in the 2011 Political Declaration on the Prevention and Control of NCDs [2]. The Sustainable Development Goals (SDGs) also include a target to reduce the proportion of premature deaths due to NCDs, including DM, by one third by 2030, with new commitments made at the 74th World Health Assembly [3] to strengthen efforts around prevention and control, as well as a set of global targets for diabetes coverage. In 2021, approximately 537 million people had diabetes worldwide, a number which is estimated to rise to 700 million by 2045 [4]. Type 1 Diabetes Mellitus (T1DM) is one of the most common endocrine and metabolic conditions in childhood and adolescence. The European Region has the highest number of children and adolescents living with T1DM compared to any other global region, reaching about 295,000 in total [4]. Wide variation in incidence of T1DM in children younger than 15 years is described by registry reports from the EURODIAB study group within Europe [5], which also indicated an overall pooled rate of annual increases of 3.4% (95% Confidence interval 2.8%–3.9%).

Until recently, type 2 Diabetes Mellitus (T2DM) was referred to as “adult-onset diabetes,” reflecting it as a disorder of ageing. Yet its prevalence is also increasing in children and adolescents, albeit highly variable among countries. According to the literature, the incidence of T2DM between the ages of 10 and 19 years in the United States was 12.5 cases per 100,000 in 2011–2012, while in the United Kingdom and other European countries, its incidence has been much lower, at <1 case per 100,000 [6], although higher in specific ethnic minorities. In 2018, across the European Union (EU), almost one in five 15 year-olds was either overweight or obese [7]. The rise in prevalence of overweight or obesity [8], which are major risk factors for T2DM, is thought to have contributed to rising rates of T2DM in adolescents [9]. Indeed, screening studies have reported the prevalence of T2DM of 0.4% up to 1% in obese adolescents ≥12 years [10]. The rise of T2DM among adolescents is a major public health concern, as earlier onset of DM implies a longer lifetime exposure to hyperglycaemia, and a consequent larger predisposition to long-term complications (e.g., renal failure, cardiovascular morbidity, retinopathy, neuropathy), premature death, and poor quality of life [11].

To our knowledge, despite the available population studies, no comprehensive analysis is yet available in Europe on both types of DM, and reliable data on adolescents are still sparse [12]. Indeed, considering their effects on health and quality of life, and the substantial direct and indirect economic burden [12], investigation of the burden of T1DM and T2DM in adolescents in European countries is overdue. In this regard, Western Europe is an important sub-region to explore the burden and epidemiological trends of T1DM and T2DM among adolescents, given the considerable epidemiological similarities and homogeneity among countries. Given the inconsistency and sparsity in data collection systems and the major data gaps for adolescents [4], we used estimates provided by the Global Burden of Diseases, Injuries, and Risk Factors study 2019 (GBD 2019) to: 1) describe the prevalence, incidence, and disability burden of T1DM and T2DM among 10–24 year old adolescents by age, sex, and country in 2019; 2) assess the disease burden of T1DM and T2DM through temporal trends using prevalence, incidence and years lived with disability (YLDs), from 1990 to 2019 at sub-regional level and national level; 3) estimate the burden of T2DM in YLDs attributable to high body mass index (HBMI).

This study adopted the broad age definition for adolescence from 10 to 24 years because it accurately captures the biological, social, and neurocognitive development of this population [13]. Additionally, it used the GBD 2019 country grouping for Western Europe, which comprised 24 countries, including Israel, that were grouped based on epidemiological similarities and/or geographical proximity.

In 2019, there were 320,237 (95% UI 234,473–408,257) adolescents aged 10–24 years old in the 24 Western European countries with T1DM, and 227,299 (130,968–345,987) with T2DM. Age groups disaggregation are reported in Supplementary Tables S1, S2.

This study is the first systematic analysis of the incidence, prevalence, and disability burden of both types of DM among adolescents in Western Europe using GBD 2019 estimates. Overall, our findings highlight the increase in incidence, prevalence, and disability burden of T1DM and T2DM. All 24 countries reported a rising trend of the three measures in T2DM over the 30 years from 1990 to 2019. This was very similar for T1DM, in that 22 of 24 countries reported increases in prevalence and YLD rates, and 23 countries reported an increase in incidence. While T1DM remains the predominant type of DM in adolescents in Western Europe, there is a striking increase in the incidence of T2DM [19], showing that this disease should not only be considered an adult-onset disease. At the very least, one of the reasons for an expected increase in prevalence of T1DM is the decline in mortality that accompanies higher quality healthcare resulting in greater survival into adolescence from incident T1DM in earlier childhood.

The rising prevalence of adolescent DM also presents new challenges for health systems. Innovative approaches are needed as the best outcomes for both types of DM require high levels of self-care by adolescents which is best supported by patient-centred care and multi-disciplinary models of care that span primary and specialist settings. A paradigm shift is required for many paediatric diabetes services which predominantly focus on T1DM and are dominated by technological advances that are less relevant for those with T2DM. Most older adolescents will be managed by adult-oriented health services. While these are typically more concerned in addressing T2DM, they are also historically oriented to older adults than adolescents. Puberty is associated with metabolic changes that of itself are thought to be associated with greater insulin resistance [20]. These biological changes intersect with the wider aspects of psychosocial development that define adolescence, with poorer adherence to treatment, initiation of risky behaviours, incident eating and mental health disorders, which are associated with complications of DM [20]. Greater self-care is required by adolescents, which is supported by carers who are equally responsible for shaping their children’s health behaviours and supervising medical management. Negotiating these changing roles and responsibilities can impose considerable challenges for individuals and their families, as well as healthcare teams [20]. Given that the transfer from paediatric to adult services typically occurs between 15 and 19 years of age [21], a further challenge for diabetes care results from the challenge of transferring specialist healthcare from paediatric oriented to adult providers [20]. Improving the health-related quality of life for adolescents is an important priority during the transition from paediatric to adult healthcare [22], which beyond the medical supervision of insulin regimens requires attention to diabetes education, social support, and psychological services. Educating and training the current and future health workforce about how to work with adolescent patients and their families is globally indicated [21], as well as in Europe [23].

In line with other studies, our findings also highlight the higher prevalence of T2DM and higher HBMI attributable YLD for T2DM among female adolescents. These sex differences might be linked to differential sex-hormone effects, undiagnosed polycystic ovary syndrome, divergent post pubertal weight gain and behavioural patterns [20]. For the latter, as sex differences in some adolescent health behaviors, such as physical activity, have been related to societal gender inequality, a society-level perspective should be adopted to address gender norms and disparities [24].

HBMI is a major risk factor for T2DM, and studies report that over 85% of children with T2DM are either overweight or obese at diagnosis [25]. The strong inverse association between BMI and age at diagnosis of T2DM also suggests an association between the duration of obesity and the risk of T2DM, which underscores the importance of preventing obesity in early childhood [26]. Adolescence has also been shown to be a highly dynamic period for weight gain, including incident obesity, and there is an association of weight gain during puberty and the risk of T2DM [13, 25]. The significance of these developmental periods reinforces the value of preventive interventions that take a life-course perspective to preventing overweight and obesity in young children as well as adolescents [26]. While most Western European countries have implemented obesity prevention policies, the strong relationship between the increasing rate of obesity in recent decades and the increasing incidence of T2DM among adolescents suggests much greater effort is indicated to slow the rising burden of DM across Europe [25]. In this context, DM among adolescents should be included within a social and commercial determinants of health framework, which will require a more comprehensive agenda aimed at the individual and societal level, as well as organizational and policy level. Greater emphasis may be warranted to ensure that preventive strategies focus on adolescents from lower socio-economic status (SES) and ethnic minorities as these are disproportionately affected by both T2DM and obesity [11] which are likely to have been exacerbated by the socio-economic, education and health inequalities caused by the COVID-19 pandemic [27].

These results show that while most countries have established DM strategies and national plans, a major barrier in the implementation of health policies is the absence of national registries. This averts the prioritisation of interventions which could be achieved through better understanding of trends in both forms of DM, and their associated complications. The lack of national registries is not restricted to T1DM, but it is this form of DM where the relevance of registries is most apparent, given the historic burden of complications that includes disabling sequelae and premature death, and the relevance of quality health services to reducing complications. The variability of estimates produced by different sources makes the absence of national registries immediately apparent. Although the global number of people with DM reported by GBD 2019 (463 million in 2019), the IDF 9th Atlas (460 million in 2019) and the NCD-Risk Collaboration (422 million in 2014) are aligned [15], greater variability for T1DM is reported [28]. GBD 2019 reported a prevalence of 1.4 million and 2.7 million cases of T1DM for 0–14 year-olds and 0–19 year-olds, while the IDF reported less than half the number, at 651,700 and 1.2 million respectively [28]. While this degree of variability will reflect differences in methods, case inclusion criteria and data sources used, it primarily reflects problems in primary data availability and quality; preliminary findings from the WHO Regional Office for Europe show that only two of the 24 Western European countries (Denmark and Sweden) included in our study have national diabetes registries. Ten countries have no registries, nine have registries for certain age groups or types of diabetes, and the situation was unknown in a further three countries [29]. Beyond disease surveillance, diabetes registries can be used to monitor aspects of clinical management, healthcare quality and healthcare costs, which are critical to many different aspects of governance and accountability, including policy implementation. Within and across countries, data harmonisation and standardisation should be prioritised as a strategy to enhance quality healthcare and patient outcomes, as well as a means of providing more precise knowledge of the direct and indirect economic burden of DM to the health system and society, which are estimated at about USD 189 billion in 2021 in the EUR Region [4].

This study shares the limitations of GBD 2019, which have been described in detail elsewhere [16, 19] Our analysis has several limitations related to the variation in the availability and quality of primary data for adolescents in general, including the paucity of data for 10–14 year-olds, and the particular paucity of DM data for 15–19 year-olds, thus there should be caution in interpretation [30–32]. The estimates of disease burden are surrounded by considerable uncertainties. Indeed, not all sources of uncertainty could be routinely captured in either the epidemiological or cause-of-death modelling processes. Differences in data availability among countries can generate difficulties in the interpretation of comparisons. Moreover, the wide uncertainty intervals for T2DM indicate caution in interpreting comparisons across countries. In this context, the addition of new and better-quality data will be essential to generate more reliable estimates in the future. Another limit is the unavailability of estimates for T2DM in adolescents <15 years old, as GBD uses the assumption that T2DM is >15 years old. This should be imputed to limited population-based data available for adolescents <15 years old. Additionally, considering that the PAF for risk factor-disease pairs are exclusively calculated for ages where both the risk and the disease are estimated, only high fasting plasma glucose and HBMI have been retrieved as risk factors for DM in the study population. However, as the proportion of T1DM and T2DM burden that is attributable to high fasting plasma glucose is always 100%, this metabolic risk factor has not been reported. HBMI was also only able to be analysed for the population aged 20–24 years old, due to the age restrictions reported above, notwithstanding that in terms of exposure values, all age groups <24 years showed an increase in HBMI in the 30 year period, finally leading to an increase in T2DM later in life (Supplementary Figures S7, S8). Other potentially important limitations relate to the lack of data on ethnicity and socio-economic status which are other risk factors for DM. Despite these limitations, a strength of this study is that it provides a comprehensive picture of the burden of DM across 24 Western European countries which will hopefully provide a benchmark for policymakers in regard to where investments are best prioritized.

In conclusion, this study describes striking increases in the incidence, prevalence, and disability burden of T1DM and T2DM among adolescents in Western Europe over the past 30 years. T1DM requires quality health system responses, which as reported in a recent scoping review [33] should expand implementation strategies, financial arrangements, and patient‐reported outcomes. For T2DM, policy is urgently needed on primary prevention as well as health system responses [34]. These should be based on understanding the relative impacts of the social and commercial determinants of health across the life course and extend beyond health service responses. Both forms of DM need further research to better understand their pathophysiology among adolescents and the interaction of genetics, puberty, and the environment. A specific focus on vulnerable populations appears indicated given the anticipated effect of intersectionalities (e.g., gender, ethnicity, low SES) on health outcomes, including mortality, as better appreciated across the COVID-19 pandemic.