In This Issue of Diabetes Care

Around one-quarter of juvenile type 1 diabetes diagnoses in Germany were accompanied by episodes of diabetic ketoacidosis (DKA), according to an analysis by Auzanneau et al. (p. 1807). Crucially, however, the frequency of DKA was not uniform across the country, with rates lowest in least deprived areas but that grew with increasing social deprivation. Rural areas also had significantly higher DKA rates than more urbanized areas. Based on the prospective Diabetes Prospective Follow-up Registry (DPV), the study focuses on newly diagnosed type 1 diabetes cases in children and adolescents (i.e., aged ≤18 years) from across Germany in the period 2016–2019. Individuals were assigned to quintiles of regional socioeconomic deprivation and to a degree of urbanization according to their postal codes. With regression modeling, the authors then looked at frequency of DKA and any association with socioeconomic deprivation and urbanization while accounting for a range of confounding factors. They found that frequency of DKA in newly diagnosed cases was lowest in the least deprived regions at ∼21% and highest in the most deprived regions at ∼27%, corresponding to a relative increase in the DKA rate of about 30%. Rural areas also had a significantly higher rate (∼28%) than towns and suburbs (∼23%) and cities (∼24%). Notably, rural areas were more frequently classified as deprived. Adjusting for other factors (e.g., age, sex, or migration background) did not change the results. “The increase of DKA at pediatric diagnosis of type 1 diabetes in recent years in many high-income countries is worrying and unexplained,” said author Marie Auzanneau. “In addition to this continuous increase observed both in the U.S. and Europe, numerous studies found even higher DKA rates during the COVID-19 crisis. Our data from a country with complete coverage of health insurance point toward delayed access to health care and/or misdiagnosis in rural and deprived areas. These results may have wide-reaching implications for prevention initiatives, either by education or antibody screening.”

Mortality due to diabetes in 2014 in areas of Seattle that were subject to the 1930s practice of redlining was nearly 54% higher in areas historically rated “hazardous” or given a grade of “D” compared with the areas rated “best” or “A.” Redlining was a practice of the long-gone Home Owners’ Loan Corporation (HOLC) that systematically discriminated against Black communities for the purposes of mortgage approvals and other financial arrangements in the 1930s and 1940s. The practice became a byline for inequality and discrimination, and according to the analysis by Linde et al. (p. 1772), its effects are still being felt, particularly in terms of diabetes health outcomes, at least up until 2014. The authors report that HOLC redlining in Seattle explains 45–56% of the variation in diabetes mortality between the years 1990 and 2014. Over the same period, the policy also explained 51–60% of the variation in years of lost life due to diabetes. HOLC originally graded residential areas of Seattle A–D, with D typically used for communities with a majority Black population. According to the authors, in 2014, a unit-higher historic HOLC score (A–D) was associated with nearly 54% higher diabetes mortality rates and nearly 67% higher years of lost life. Redlining was outlawed in the late 1960s. Commenting more widely, author Leonard E. Egede told Diabetes Care: “While significant attention has been brought to the issue of social determinants and social risk factors as drivers of poor health outcomes for diabetes, there is limited appreciation of the fact that structural racism and/or structural inequalities are antecedent to social determinants of health. Therefore, this study provides strong evidence for the ongoing impact of structural racism and historic redlining on health outcomes for diabetes. Next steps are large population-based studies that elucidate the mechanisms and pathways that explain this effect as a starting point for designing interventions to address the problem.” For a wider mapped view of the issue of historic redlining in the U.S., see https://bit.ly/3xz4jYN. The Seattle map is available top right.

Intake of high levels of dietary protein appears to increase type 2 diabetes risks, according to Li et al. (p. 1742). The amount and type of protein, however, may modify risks, while swapping certain dietary protein sources may decrease risks. Based on the findings, the authors suggest attention should be given to dietary protein sources as a method to modify the risks for type 2 diabetes. The findings come from a primary cohort of just under 109,000 women without diabetes at baseline who are part of the Women’s Health Initiative. In addition, nearly 35,000 adults without diabetes from the UK Biobank were also included as a replication cohort. The authors then used modeling with adjustment for a large set of factors to estimate protein–type 2 diabetes risk associations. They found that diabetes risks were elevated with increasing consumption of protein from total, animal, red meat, processed meat, poultry, eggs, and low–omega 3 fish sources. In contrast, protein from plants, whole grains, and nuts reduced risks. Sources that were risk-neutral included high–omega 3 seafood, various dairy sources, and legumes. Notably, however, nearly all the associations were attenuated when the models were adjusted for BMI and waist-to-hip ratio. In further analyses, the authors looked at the effect of substituting 5% of energy from animal protein sources with various plant, dairy, or fish protein sources. Nearly all substitutions resulted in some reduction in risks. Finally, case-control studies nested within the primary cohort identified a series of biomarkers relating to obesity and inflammation that may mediate the protein–diabetes risk associations. Again, adjustment for BMI attenuated the effects. Commenting further on the study, author Simin Liu said: “Substituting animal protein sources with plant protein sources may decrease risk of type 2 diabetes by reducing adiposity-related inflammatory cytokines. These findings support the notion that the quality and sources of dietary protein are important for the prevention of type 2 diabetes.”

Achieving normal birth weight can be challenging in the context of type 1 diabetes with pregnancies often ending with large for gestational age (LGA) infants. According to Scott et al. (p. 1724), the key to achieving normal-weight births is continuous glucose monitoring (CGM). Specifically, they suggest that achieving significantly lower glucose concentrations over 24 h and higher time in range before the end of the first trimester should be key aims for reducing the chances of LGA births. According to the authors, traditional HbA1c or self-monitored capillary glucose measures often do not give the insights of CGM, which can give a user hundreds of measurements per day and thus inform diabetes self-management. However, international guidelines on CGM do not give guidance on “gestationally appropriate” glucose targets, motivating the study. Based on >10.5 million CGM measurements from 386 pregnant women with type 1 diabetes, the study looked at various glucose metrics and profiles to determine features associated with normal and LGA births. The authors found that CGM-measured glucose levels fell as pregnancy progressed. Time in range (3.5–7.8 mmol/L) also increased up to 10 weeks and then remained stable up to 28 weeks before further improvements up to delivery. Crucially, key glucose metrics diverged at 10 weeks’ gestation according to whether final birth weights were normal or LGA. Women who had normal births had significantly lower glucose levels and higher percentage of time in range compared with those who had an LGA birth. In the latter case, 24-h maternal glucose profiles were consistently higher with LGA births than with normal births. Commenting further, author Eleanor M. Scott said: “As increasing numbers of women use CGM instead of fingerprick to manage their diabetes during pregnancy, it is important for us to change from fingerprick glucose targets and instead focus on the CGM metrics to aim for at each stage of pregnancy. The 24-h profile data show that good mealtime glucose management remains important to achieve these CGM metrics.”