Investigators found that there was no association of dementia incidence in insulin use at 5 years after stopping metformin and an association at 0.07 years for HbA1C level at year 1.
Metformin cessation has been associated with an increased dementia incidence in a new study published on JAMA Network Open. Investigators believe that the results are important for the clinical treatment of adults with diabetes.
The study investigators aimed to determine if there was an association between stopping metformin, for a reason unrelated to kidney dysfunction, and dementia incidence. They stated that other studies have suggested that metformin could be associated with a reduced dementia incidence; however, this was affected by disease severity and prescribing trends, according to the study authors. They added that the discontinuation of metformin for those with diabetes is typically due to kidney dysfunction, though there are incidences where it is due to a less serious adverse effect.
In the study, investigators included individuals from Kaiser Permanente Northern California. The patients were born prior to 1955 without a history of diagnosed kidney disease before beginning metformin. According to the study authors, dementia follow-up started in 1996, with the implementation of electronic health records, and continued to 2020. The data were analyzed from November 2021 through September 2023.
The main outcome was all-cause incidence of dementia. Investigators also measured hemoglobin A1C (HbA1C) levels and insulin usage at 1 and 5 years after metformin was discontinued. They used this to determine if the changes in blood glucose or insulin contributed to the association between stopping metformin and dementia incidence.
The study authors reported that a total of 12,220 participants discontinued metformin early and had normal estimated glomerular filtration rate (eGFR). They were compared to metformin users, who had either not stopped treatment or had stopped treatment, with or without restarting, after their first abnormal eGFR measurement. Individuals in both groups were matched by age, gender, and duration of diabetes.
Of the 12,220 individuals, 46.2% were women, with a median age of 59.4 years at the start of their first metformin prescription. There were also 29,126 individuals in the metformin user group, with 46.6% being women and a mean of 61.1 years when starting their first metformin prescription.
Individuals who stopped taking metformin early had a 1.21 times greater hazard of dementia diagnosis compared with those who routinely used metformin. Furthermore, investigators found that there were no association in insulin use at 5 years after stopping metformin and 0.07 years for HbA1C level at year 1 after stopping treatment. The study authors concluded that the association of dementia and metformin cessation was independent of changed in HbA1C level and insulin usage.
The study authors said the results of this study could have potential implications for future diabetes treatment, particularly for treatment later in life. Additionally, they said that finding ways to manage or mitigate gastrointestinal AEs of metformin might be beneficial for those at high risk of dementia, including those with the APOE ε4 or those who have a history of dementia in their family.
The investigators noted limitations of the study, which included that the dementia diagnoses were obtained based on medical record, which they said follow the onset of pathology. Furthermore, the study was a case analysis, therefore the main analysis and mediation samples were different sizes, according to the study authors.
The study authors concluded that the results demonstrated the association between increased dementia incidence and metformin cessation, supporting previous research indicating that starting metformin was associated with a reduced risk of dementia.
Zimmerman SC, Ferguson EL, Choudhary V, Ranatunga DK, et al. Metformin Cessation and Dementia Incidence. JAMA Netw Open. 2023;6(10):e2339723. doi:10.1001/jamanetworkopen.2023.39723