Each year of delay in type 1 onset is linked to lower mortality

Design

Retrospective nationwide register cohort

Population

34,155 adults with type 1 diabetes and 150,026 population comparators, Sweden

Primary outcome

All-cause mortality, cardiovascular and kidney outcomes by age at diagnosis

Effect

HR 0.98 (95% CI 0.98 to 0.99) per 1-year increase in age at diagnosis; HR 3.23 (2.92 to 3.58) for onset at 0 to 10 years

Using the Swedish National Diabetes Register, researchers compared 34,155 adults with type 1 diabetes against 150,026 comparators from the population register, following them until the end of 2020.

The hazard ratio per additional year of age at diagnosis was 0.98 (0.98 to 0.99) for all-cause mortality, which the authors express as about 8.2% lower risk per five years. For cardiovascular mortality it was 0.97, and for non-cardiovascular mortality 0.99. Excess risk against the general population was greatest in those diagnosed between 0 and 10 years of age: HR 3.23 for all-cause mortality and 5.73 for cardiovascular mortality.

In practice this identifies people diagnosed in early childhood as the group with the longest exposure and the highest relative cardiovascular risk. It supports early, sustained attention to lipids, blood pressure and kidney screening in that group, within existing guidance.

• Diagnosis before age 10 carried an all-cause mortality HR of 3.23 against the general population.
• Cardiovascular mortality excess was 5.73 times higher in those diagnosed before age 10.
• Each additional year of age at diagnosis was associated with 2% to 3% lower cardiovascular mortality.
• Note the age at onset in the record, because it adds to risk stratification.
• Existing screening schedules still apply: this does not change the guideline.

Why it matters

Age at onset is a cheap, already-recorded marker that tracks long-term cardiovascular risk.

Don't overread it

A register cohort shows association only; duration of exposure and calendar period are tangled with age at onset.

The statistics, in plain English

A hazard ratio of 0.98 per year looks tiny, but it compounds across years: a person diagnosed ten years later has roughly 17% lower hazard on this estimate. The very narrow confidence intervals reflect the size of the register, not certainty about the cause.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.