Home / Resources / Articles / Metformin Linked to Increased Risk of Dementia and Parkinson’s Disease

Metformin Linked to Increased Risk of Dementia and Parkinson’s Disease

Apr 29, 2017

Study finds connection between duration of therapy in senior patients and development of neurodegenerative disease.

In 2011, the Journal of Alzheimer’s Disease published the findings of a large Taiwanese study showing a protective effect against development of dementia in diabetes patients who were given oral antidiabetic agents. The cohort of over 100,000 subjects included patients over 50 with type 2 diabetes, who were free of dementia at initiation, and received either or both metformin and a sulfonylurea. The results suggested that while T2D carries a two-fold increase in the risk of dementia, use of metformin, sulfonylureas, or both can reduce the risk by up to 35% over eight years. Medscape recently reported that at AD/PD 2017 (the 13th International Conference on Alzheimer’s and Parkinson’s Diseases), a group of Taiwanese neurologists presented the results of their own study looking at possible risk increases for Alzheimer’s and Parkinson’s in people with type 2 diabetes, citing uncertainty about the effects of metformin on the risk of developing neurodegenerative diseases.

Utilizing the National Health Insurance research database in Taiwan, 4,651 diabetes patients who had prescriptions for metformin were selected, along with an identically sized, matched control group of patients not taking metformin. Although the statistical methods were not made available at the time of this writing, the findings were rather striking. Patients were retrospectively followed over a 12-year period. With regards to Parkinson’s disease, the event rate in metformin users was 6.85% (n=318) vs. non-users at 2.78% (n=129; HR 2.27; 95% CI 1.66-3.07), after adjustment for age, gender, and severity of diabetes.  All-cause dementia showed 11.5% (n=534) vs 6.7% (n=311; HR 1.66; 95% CI 1.35-2.04), Alzheimer’s dementia 1.64% (n=76) vs 0.83% (n=38; HR 2.13; 95% CI 1.20-3.79), and vascular dementia 1.64% (n=76) vs 0.69% (n=32; HR 2.30; 95% confidence interval 1.25-4.22).

Further examination revealed a connection between duration of therapy and development of neurodegenerative disease, with an absolute occurrence difference for Parkinson’s favoring metformin users by 5.9% (HR 1.77; 95% CI 1.17-2.68) for durations of therapy less than 180 days, all the way up to 14.3% (HR 4.49; CI 3.06-6.58) for metformin use at or above 400 days. The same comparison made for all-cause dementia also favored metformin users by 7.99% (HR 1.02; 95% CI 0.74-1.41, ns) to 20.6% (HR 2.84; 95% CI 2.12-3.82) over the same < 180 days to ≥ 400 days duration range. Cumulative dosage also had a similar increase in risk of Parkinson’s, with annual dosages ranging from < 130 grams metformin at 5.35% (HR 1.58; 95% 1.02-2.44) to > 385 grams at 9.38% (HR 3.54; 95% CI 2.41-5.20). For all-cause dementia, the increase in risk ranged from 9.97% for < 130 grams annually (HR 1.22; 95% CI 0.90-1.67, ns) to 12.3% for 240-385 grams annually (HR 2.06; 95% CI 1.54-2.75). Interestingly, in patients taking more than 385 grams of metformin per year, the risk of all-cause dementia slightly decreased to 11.9% (HR 1.97; 95% CI 1.45-2.68). While patient characteristics were not available in this report, it can only be assumed that the number of patients in each dosing range quartile was roughly the same (i.e. n~1162 in each group). Conversely, the group in ≥ 400 days duration arm could have been on metformin for the entirety of the 12-year period, but again, this was not provided.

The presenting investigator did admit limitations to their study, including the statement that they were not experts in the field of diabetes management, but did suggest that other studies be performed looking at incident neurogenic diseases with other diabetes treatment modes, including sulfonylureas and insulin. She also stated that while their findings could not be considered conclusive, that “clinicians should keep them in mind.”  Dr. Larry Ereshefsky of the Follow the Molecule group stated that the findings of this study, while interesting, cannot be considered applicable, as the investigators did not report how they controlled for confounders, nor did they report the effects of other treatments these patients were receiving or how effectively their diabetes was being controlled. Looking at the provided statistics, there are also three data points suggesting a protective effect from metformin, as demonstrated by reported confidence intervals.  In reporting the risk of Parkinson’s, in the group receiving metformin for 180-300 days, the confidence interval for hazard ratio crossed zero (95% CI 0.90-2.37). For all-cause dementia, two groups had the same issue: annual doses < 130 grams (95% CI 0.90-1.67) and duration of metformin < 180 days (95% CI 0.74-1.41).  This shows no statistical significance in these groups, and suggests there may be a portion of patients who actually have reduced risk of neurogenerative diseases when given metformin.  Future studies of better design will be necessary to support these claims.

Practice Pearls:

  • Type 2 diabetes has long been understood to carry an increased risk of developing neurodegenerative diseases.
  • Until recently, metformin and sulfonylurea use in T2D has been associated with a decrease in risk for Parkinson’s and Alzheimer’s diseases.
  • A recent study suggests metformin use can increase the risk of Parkinson’s and Alzheimer’s, but there are also concerns over the validity of the findings due to study design flaws.


Hsu CC, Wahlqvist ML, Lee MS, Tsai HN. Incidence of dementia is increased in type 2 diabetes and reduced by the use of sulfonylureas and metformin. J Alzheimers Dis. 2011;24(3):485-93.

Metformin Use Linked to Increased Dementia, Parkinson’s Risk in Patients with Diabetes. Medscape. Mar 30, 2017.

“Diabetes – A Major Risk Factor for Kidney Disease.” The National Kidney Foundation. N.p., 03 Feb. 2017. Web.

Penno, Giuseppe, Eleonora Russo, Monia Garofolo, Giuseppe Daniele, Daniela Lucchesi, Laura Giusti, Veronica Sancho Bornez, Cristina Bianchi, Angela Dardano, Roberto Miccoli, and Stefano Del Prato. “Evidence for two distinct phenotypes of chronic kidney disease in individuals with type 1 diabetes mellitus.” Diabetologia (2017). Web.
Mark T. Lawrence, University of Colorado-Denver, School of Pharmacy NTPD