Personalized diet and medications for cognition – Garcia-Ptacek group

The research group develops personalized medication and dietary interventions to prevent and treat dementia, with a focus on medication repurposing to identify new therapeutic options. Machine learning is applied both to support repurposing efforts and to improve predictions of cognitive decline. In parallel, the group works on methodological development to refine our epidemiological approaches and improve causal inference in our cohort studies.

Publications

All publications from group members

Funding

Grants

  • Medication repurposing in dementia: selecting brain-friendly medications and establishing a new indication
    Center for Innovative Medicine
    1 January 2026 - 31 December 2028
  • Medication repurposing to slow cognitive decline in Alzheimer's dementia
    ALF Medicin-Region Stockholm
    1 January 2026 - 31 December 2028
  • Neuroprotection with Repurposed Cardiovascular Drugs in Alzheimer's Disease
    Alzheimerfonden
    1 January 2025 - 31 December 2025
  • Innovative ways to fight Alzheimer´s disease: Fluid biomarkers and intracellular therapies
    Private donation from Leif Lundblad Family and others
    1 October 2023 - 30 September 2026
  • Swedish Research Council
    1 January 2023 - 31 December 2026
    Dementia creates a great personal and societal burden and there are currently no treatments to stop memory loss.  Many patients with dementia take medications to treat other conditions, eg. high blood pressure and cholesterol. Some of these medications may have central effects on the pathophysiological processes leading to dementia.Medication repurposing is a cost and time-effective way to discover new treatments. Swedish registers are a unique tool to detect medication candidates for repurposing. The Swedish Dementia Registry (SveDem) has &gt
    100000 patients and ca. 144000 measures of cognition over time (measured with the Mini-Mental State Examination—MMSE), making it perfect to examine whether certain medications are associated with less cognitive decline in patients with dementia. Cholesterol and kidney function affect treatment and cognition and are available for 25000 patients. We propose a study combining traditional cohort studies and newer artificial intelligence (machine learning) techniques to detect medications associated with slower cognitive decline in patients from SveDem. We will test pre-specified hypotheses since we suspect specific medications are better candidates to prevent cognitive decline, but we will also let the machine learning algorithms explore the data to find associations that we do not suspect in advance. Finally, the candidate medications will be tested in animal and cellular models to determine the mechanisms of their effects.
  • Translational identification of medication repurposing candidates for Alzheimer's dementia
    Swedish Research Council
    1 January 2023 - 1 January 2026
    Dementia creates a great personal and societal burden and there are currently no treatments to stop memory loss. Many patients with dementia take medications to treat other conditions, eg. high blood pressure and cholesterol. Some of these medications may have central effects on the pathophysiological processes leading to dementia. Medication repurposing is a cost and time-effective way to discover new treatments. Swedish registers are a unique tool to detect medication candidates for repurposing. The Swedish Dementia Registry (SveDem) has >100000 patients and ca. 144000 measures of cognition over time (measured with the Mini-Mental State Examination—MMSE), making it perfect to examine whether certain medications are associated with less cognitive decline in patients with dementia. Cholesterol and kidney function affect treatment and cognition and are available for 25000 patients. We propose a study combining traditional cohort studies and newer artificial intelligence (machine learning) techniques to detect medications associated with slower cognitive decline in patients from SveDem. We will test pre-specified hypotheses since we suspect specific medications are better candidates to prevent cognitive decline, but we will also let the machine learning algorithms explore the data to find associations that we do not suspect in advance. Finally, the candidate medications will be tested in animal and cellular models to determine the mechanisms of their effects.
  • Biomarkers and Intracellular treatments
    New Innovative Ways to Treat Alzheimer—Leif Lundblad and others-project
    1 January 2023 - 30 September 2025

Staff and contact

Group leader

All members of the group

Visiting address

Karolinska Institutet, Institutionen för Neurobiologi, Vårdvetenskap och Samhälle (NVS), Division of Clinical Geriatrics, Blickagången 16, 7e våning, Huddinge, Stockholm, 141 52, Sweden; Karolinska Institutet, NVS. Sektionen för klinisk geriatri/7e, Huddinge, Stockholm, 14183, Sweden

Postal address

Karolinska Institutet, NVS. Clinical Geriatrics, Huddinge, Stockholm, 141 83, Sweden

Keywords:
Diet, Food, and Nutrition Drug Repositioning Elder Nutritional Physiological Phenomena Epidemiology Geriatrics Show all
Sara Hedberg
29-09-2025