Higher Risk of Dementia Linked to Earlier Onset of Diabetes


New data find significant associations between every five-year earlier onset of diabetes and a higher hazard rate of dementia, in patients aged 35-75 years old.

Archana Singh-Manoux, PhD

Archana Singh-Manoux, PhD

A recent study found patients with a younger age at onset of diabetes had a significant association with higher risk of subsequent dementia.

Investigators, led by Archana Singh-Manoux, PhD, of EpiAgeing at the Université de Paris, found for every five-year earlier onset of diabetes, there was a significant association with higher hazard of dementia.

The team assessed the data in the Whitehall II ongoing cohort study, a collection of 10,308 people employed in London government departments, established in 1985–1988. Follow-up examinations take place every 4-5 years, with 2015–2016 being the most recent completed data and 2020–2021 still being collected.

Investigators also linked the data to the UK National Health Service (NHS), with records updated annually until March 2019.


The study included 7 clinical assessments between 1985 and 2016 where venous blood samples were taken either in the morning after a ≥8 hour fast or a ≥5 hour fast after a light, fat-free breakfast. Fasting glucose was measured with the glucose-oxidase method.

Each examination collected data on self-reported physician-diagnosed type 2 diabetes, as well as prescriptions of diabetes medications. The national Hospital Episode Statistics (HES) database was linked for follow-up on participants.

Investigators defined Type 2 diabetes with measures of fasting glucose (126 mg/dL) at the clinical examination, use of medication, physician diagnosed diabetes, or record of diabetes in the HES database from 1985 – 2019.

The team’s secondary exposures included the Finnish Diabetes Risk Score (FINDRISC), fasting glucose, and prediabetes.

Dementia cases were assessed through linkage to 3 registers, including the HES, the Mental Health Services Data Set, and the Office for National Statistics Mortality Register, until March 2019.


Within the study period, 639 patients (6.3%) were diagnosed with dementia. Investigators said they were more likely to have chronic conditions during the follow-up period and have a worse cardiovascular risk profile at 60 years old.

In the follow-up, 1710 participants (16.9%) developed diabetes and 153 (8.9%) of these participants were subsequently diagnosed with dementia.

At age 55, the dementia rate per 1000 person-years was 3.14 in participants without diabetes and 5.06 in participants with diabetes, at a hazard ratio (HR) of 2.14 (95% CI, 1.44 – 3.17).

At age 70, the dementia rate was 8.85 per 1000 person-years in participants without diabetes and 13.88 per 1000 person-years in participants with diabetes, with a HR of 1.58 (95% CI, 1.22 – 2.03).

Diabetes cases in patients ≥70 years included 48.3% of cases. Investigators treated diabetes and covariates as time-varying measures. The dementia rates were 1.76 per 1000 person-years in participants without diabetes and 6.25 in participants with diabetes, with a HR of 1.40 (95% CI, 1.15-1.70).

The data for diabetes onset from 0 - 5 years earlier had significant association with subsequent dementia with an HR of 1.53 (95% CI, 1.03-2.29) and a dementia rate of 8.63 per 1000 person-years.

Diabetes onset from 6 - 10 years earlier had a dementia rate of 10.51 per 1000 person-years.

Investigators considered diabetes status at age 70, where the HR of patients with diabetes onset more than 10 years earlier was 2.12 (95% CI, 1.50-3.00), with a dementia rate of 18.30 per 1000 person-years.

Diabetes onset 6-10 years earlier had an HR of 1.49 (95% CI, 0.95-2.32), with a dementia rate of 12.99 per 1000 person-years.

The rate of diabetes onset 0-5 years earlier had an HR of 1.11 (95% CI, 0.70- 1.76), with a dementia rate of 10.00 per 1000 person-years.

In the analysis adjusted for sociodemographic and health factors, the team found that at age 70, every 5-year younger age at the onset of type 2 diabetes had an association with an HR of dementia of 1.24 (95% CI, 1.06- 1.46).

The team found little association between dementia and either the FINDRISC or preclinical diabetes in participants without diabetes, regardless of age.


Investigators concluded younger age at onset of diabetes had a significant association with a higher HR of dementia.

“Taken together, these findings highlight the importance of age at onset of diabetes and cardiovascular comorbidity in persons with diabetes for risk of dementia,” investigators wrote.

The team did not determine the precise mechanism behind the association. They hypothesized it may be due to connections between brain metabolic dysfunction and insulin resistance, as well as higher rates of hypoglycemia in diabetes treatment increasing the risk of dementia.

“Further, studies do not always show a consistent association between diabetes and hallmarks of Alzheimer disease, such as amyloid and tau pathology,” investigators wrote. “Although the histopathological, molecular, and biochemical abnormalities in Alzheimer disease are well characterized, a unified framework that links these features is lacking.”

The study, “Association Between Age at Diabetes Onset and Subsequent Risk of Dementia,” was published online in JAMA.

Related Videos
Karen Shen, PhD: How COVID-19 Impacted Nursing Homes Staff Turnover Rates
Karen Shen, PhD: High Turnover Rates in Nursing Homes
© 2023 MJH Life Sciences

All rights reserved.