Earlier onset of type 1 diabetes tracked with higher adult mortality in Sweden

Design

Retrospective nationwide registry cohort with matched population controls

Population

34,155 adults with type 1 diabetes and 150,026 comparators, Sweden, 1998–2020

Primary outcome

All-cause, cardiovascular and non-cardiovascular mortality

Effect

HR 0.98 (0.98–0.99) per year later onset; onset age 0–10 v population HR 3.23 (2.92–3.58) all-cause

A retrospective analysis of the Swedish National Diabetes Register followed 34,155 adults with type 1 diabetes, registered between 1998 and 2019, against 150,026 matched people from the general population, published 21 September.

Each year later at diagnosis was associated with lower all-cause mortality (hazard ratio 0.98 per year, 95% CI 0.98 to 0.99) — about 8% lower per five years. The gradient was steeper for cardiovascular death (0.97 per year). Compared with the general population, people diagnosed between birth and age 10 carried about three times the risk of death from any cause (HR 3.23, 2.92 to 3.58) and nearly six times the risk of cardiovascular death (HR 5.73, 4.32 to 7.59).

The likeliest explanation is cumulative glycaemic exposure: someone diagnosed at six carries decades more of it by age 40 than someone diagnosed at 30. That makes age at onset a cardiovascular risk factor in its own right in young adults, who often score low on conventional calculators built on age.

This is observational and cannot say that intervening earlier changes the outcome, but it supports earlier attention to blood pressure, lipids and kidney markers in adults whose diabetes began in childhood.

• Record age at type 1 diabetes diagnosis as a cardiovascular risk factor in adult clinic notes.
• Consider earlier lipid, blood pressure and albuminuria review in adults diagnosed before age 10.
• Conventional risk calculators may under-estimate risk in young adults with childhood-onset diabetes.
• Use the transition from paediatric to adult care to restate long-term cardiovascular risk plainly.

Why it matters

Duration, not age, drives risk in type 1 diabetes, so a 30-year-old diagnosed at five is not a low-risk 30-year-old.

Don't overread it

This is registry data — it shows who is at higher risk, not that any specific earlier treatment lowers it.

The statistics, in plain English

A hazard ratio of 0.98 per year looks tiny, but it compounds: over five years of later onset it amounts to about 8% lower mortality. A hazard ratio of 5.73 for cardiovascular death means nearly six times the rate seen in the general population, though the wide interval (4.32 to 7.59) shows the exact figure is uncertain. All of this is association from registry data.

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.