Type 1 diabetes linked to higher mortality, with a larger relative gap in females

The study

This was a retrospective cohort study published in Diabetes Care. It used data from Maccabi Healthcare Services (MHS), a large state-mandated health care provider network. The authors included all patients with type 1 diabetes diagnosed before age 35 between 2000 and 2022. These patients were identified through a validated registry.

The comparison group had no diabetes. It was built by 1:4 propensity score matching on sex and age. Follow-up continued until 2023, death, or transfer out of MHS, whichever came first. The authors used Cox regression and sensitivity analyses.

The aim was to assess all-cause mortality in people diagnosed in recent decades, a period of major advances in diabetes care technology.

What they found

The study included 4,299 people with type 1 diabetes and 17,196 matched controls. Mean age at diagnosis was 17.4 years (SD 9.1). Overall, 45.6% were female. There were 112 deaths over 204,102 person-years.

• Mortality was 131.0 per 100,000 person-years with type 1 diabetes and 35.2 in controls.
• Type 1 diabetes was associated with higher mortality (hazard ratio [HR] 3.66; 95% CI 2.53-5.31). Estimates were consistent after adjustment and in sensitivity analyses.
• Relative excess mortality was greater in females than in males. The adjusted HR was 8.40 (95% CI 3.71-19.04) in females and 2.90 (1.78-4.72) in males (P-interaction = 0.02). Absolute mortality rates were higher in males.
• Risk was higher in people diagnosed in 2000-2010 than in 2011-2022. The adjusted HR was 4.36 (1.22-15.57) for the earlier group and 2.94 (1.31-6.60) for the later group. Follow-up was truncated to make the groups comparable.

How much weight to give it

The authors conclude that type 1 diabetes was associated with higher mortality, particularly in females, and that the association was attenuated in more recent diagnoses. This is an observational design, so it shows association and cannot show cause.

The cohort is large, and the registry is validated. However, the number of deaths is small (112). The confidence interval for the female estimate is wide (3.71-19.04), but the sex difference was statistically significant on the interaction test (P-interaction = 0.02).

The period comparison is less certain. The confidence intervals for the two diagnosis periods are wide and overlap, and the abstract reports no interaction test for them. The difference between periods should be read with caution.

The abstract does not report causes of death, HbA1c, technology use, or funding. It therefore cannot show whether newer therapies explain the lower relative risk in later years.

The relative excess was greater in females, but absolute mortality rates were higher in males. Both facts matter when interpreting the sex difference. The data come from one health care provider network (Maccabi Healthcare Services), and the abstract does not say whether the findings apply to other populations.

For your clinic

These findings support the view that people with type 1 diabetes can still face higher mortality than peers without diabetes, even among those diagnosed in recent years. The relative excess was greater in females, while absolute mortality rates were higher in males. This is a general implication, not advice for any one patient.

The abstract does not say whether the findings apply to patients in India, and it does not say which care factors drive the differences. Read the full paper for the methods, the matching details and any analysis of causes of death before changing local practice or counselling.