HealthDay News — Postmenopausal women with both diabetes and high estradiol levels were at significantly greater risk for new-onset dementia, results of a population-based study in France indicates.

Among 5,644 women aged 65 years and older who were followed for four years, those who had diabetes at study enrollment and were in the top quartile of serum estradiol had a 14-fold higher risk for new onset dementia (adjusted hazard ratio 14.2; 95% CI: 1.60-123, P=0.02) compared with those in the middle two quartiles of estradiol, Laure Carcaillon, PhD, from INSERM in Paris and Bordeaux, and colleagues reported in Neurology.

Women without diabetes, but with estradiol in the top quartile had nearly double the risk for dementia compared with those in the middle quartiles (adjusted HR 1.9, 95% CI not reported, P=0.04), the researchers found.

Continue Reading

The analysis used data from the Three-City Study, which consisted of older community-dwelling adults from Bordeaux, Dijon and Montpellier enrolled from 1999 to 2001. Participants underwent a baseline exam and then follow-up every two years thereafter for cardiovascular disease and dementia. The current analysis relied on data from the first two follow-up visits.

Patients underwent a simple screen, and those with suspected dementia were referred to a neurologist to undergo further examination and have the diagnosis confirmed based on established Alzheimer disease or vascular dementia criteria. 

Overall, 132 women developed dementia after four years of follow-up, including 90 with Alzheimer disease and 42 with vascular dementia. 

Low estradiol levels were also associated with increased dementia risk relative to middle-range levels, the researchers found. The adjusted hazard ratio for dementia onset during follow-up in the lowest estradiol quartile was 1.8 (P=0.04), the researchers found.

Among patients with diabetes, the lowest estradiol quartile had a hazard ratio for dementia of 3.4 compared with the middle quartiles; however, there were only two cases of dementia in these patients and the point estimate failed to reach statistical significance (P=0.52).

Adjusting for inflammatory and blood coagulation markers had no effect on the study findings. There was no association between testosterone levels and dementia risk.

“High estradiol level is an independent predictor of incident dementia, particularly in postmenopausal women with diabetes,” the researchers concluded. 

They added that interactions between diabetes and estradiol levels boosting dementia risk is biologically plausible, as estradiol affects vascular disease mechanisms.

Study limitations include a relatively small cohort of participants with dementia, questions about the accuracy of estradiol assays and the short follow-up period. Carcaillon and colleagues also noted that higher estradiol levels may be “a consequence of the disease that may start more than 10 years before the appearance of symptoms.”

The study researchers reported relationships with Novartis and Ipsen as well as government research grants.


  1. Carcaillon L et al. Neurology. 2014; doi:10.1212/WNL.0000000000000107.