CoChat
Sign Up
Cited by 167Diabetes Treatment and ManagementDementia and Cognitive Impairment ResearchPharmaceutical Practices and Patient Outcomes

Type 2 Diabetes and Antidiabetic Medications in Relation to Dementia Diagnosis

Cheng, Chin, Lin, Ching-Heng, Tsai, Yi-Wen, Tsai, Chia-Jui, Chou, Po-Han, Lan, Tsuo-Hung·2014·The Journals of Gerontology: Series A
DOI 10.1093/gerona/glu073
Open this paper in CoChat — free account

Abstract

BACKGROUND: Type 2 diabetes (T2D) has been shown to increase dementia risk, but few studies evaluated the relationship between antidiabetic treatment and dementia. METHODS: We followed up 67,731 participants who were nondemented, nondiabetic, aged 65 or over at baseline from January 2004 to December 2009, to observe the onset of T2D (median follow-up 2.4 years), and to compare the risk of the development of dementia associated with particular types of antidiabetic medication among participants with T2D who had solely one type of antidiabetic agents throughout the follow-up period (median follow-up for participants with T2D 3.1 years). RESULTS: The hazard ratio for dementia diagnosis in the new-onset T2D participants compared with the non-T2D participants was 1.56 (95%CI: 1.39-2.18). The relative rate of dementia was 5.31 (95% CI: 1.89-14.96) for participants taking thiazolidinediones (n = 28) and 1.22 (95% CI: 0.78-1.91) for those taking sulfonylureas (n = 796) compared to those taking metformin (n = 1,033). The risk of dementia was higher in ever (n = 841) versus never users (n = 4,579) of thiazolidinediones: 1.44 (95% CI: 1.12-1.86). CONCLUSIONS: Diabetes is associated with an increased risk of dementia. The risk effect becomes weaker provided that participants take sulfonylureas or metformin rather than thiazolidinediones for a longer period.

AI snapshot

Written from the paper’s full text.

TL;DR

In a Taiwanese population-based cohort of adults aged 65 or older, new-onset type 2 diabetes was associated with a higher risk of dementia. Among participants with type 2 diabetes, thiazolidinedione use was associated with a substantially higher dementia rate than metformin, while sulfonylurea use was not significantly different from metformin.

Key findings

  • New-onset type 2 diabetes was associated with dementia diagnosis: adjusted HR 1.56 (95% CI 1.39–2.18) versus non-diabetic participants. Dementia rate was 1.78 cases per 100 person-years in T2D participants (n = 147), 1.26 times higher than in non-T2D participants (n = 4,650).
  • Among T2D participants taking only one antidiabetic medication, thiazolidinedione users had a higher dementia rate than metformin users: relative rate 5.31 (95% CI 1.89–14.96). Sulfonylurea users did not differ significantly from metformin users: 1.22 (95% CI 0.78–1.91). At study end, dementia was diagnosed in 4/28 TZD users (14.3%), 39/1,033 MET users (3.8%), and 40/796 SU users (5.0%).
  • In a separate comparison among T2D participants, any-use of thiazolidinediones versus never-use was associated with dementia: HR 1.44 (95% CI 1.12–1.86).
  • In analysis one, hypertension (HR 1.19, 95% CI 1.13–1.25), cerebrovascular disease (HR 1.44, 95% CI 1.33–1.56), and depression (HR 2.12, 95% CI 1.88–2.39) were independently associated with higher dementia risk; hyperlipidemia was not significantly associated.

Methods

  • Population-based cohort using Taiwan’s National Health Insurance Research Database / Longitudinal Health Insurance Database 2005. The study followed 67,731 participants aged 65 or older who were dementia-free and T2D-free at baseline in January 2004 through December 2009; median follow-up was 2.4 years for T2D onset and 3.1 years for participants with T2D.
  • Four main analyses examined: (1) T2D in relation to dementia; (2) the three most common antidiabetic agents among new-onset T2D participants using solely one medication type and no insulin throughout follow-up (n = 1,857: MET n = 1,033; SUs n = 796; TZDs n = 28); (3) any TZD use (n = 841) versus never-use (n = 4,579); and (4) duration/adherence of medication use.
  • Cox proportional hazards regression models were adjusted for age, sex, hypertension, hyperlipidemia, and cerebrovascular disease; Kaplan-Meier survival curves and chi-squared or t tests were also used. Diagnoses were based on ICD-9-CM codes, with dementia defined as at least two dementia diagnosis records within any 1 year and new-onset T2D defined by outpatient/inpatient diagnosis plus oral antidiabetic medication for at least three months.

Limitations

  • The authors state selection bias is an important limitation because this was database research rather than a randomized trial; antidiabetic medication selection could be influenced by patients’ clinical condition and physicians’ preferences. They also state they lacked data on HbA1c, drinking, or smoking, and the sample was not large enough for meaningful subgroup analyses stratified by major comorbidities.
  • The authors state the follow-up period was relatively short for a chronic disease like dementia, though they argue the median 3.1-year follow-up from T2D onset was reasonable based on prior studies; longer follow-up would make associations more definitive.
  • The authors caution that results should be interpreted with caution given the observational design and the relatively small number of TZD users; in analysis two, depression was not included in the regression model because of an inadequate number of cases.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/gerona/glu073

Cites

24 of 24 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 18 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Among T2D participants, ever users of TZDs had a higher risk of dementia than never users, with an adjusted HR of 1.44 (95% CI 1.12-1.86).

    “The risk of dementia was higher in ever (n = 841) versus never users (n = 4,579) of thiazolidinediones: 1.44 (95% CI: 1.12-1.86).”
    • Also stated in: § TZD Risk for Dementia in T2D Participants (Analysis Three) · plain
      “the HR of dementia for TZD users versus nonusers was 1.44 (95% CI: 1.12-1.86)after full adjustment (Table 3).”
    Quote in the paperSubstantial claim
  • Findingown data

    In the geriatric cohort, new-onset T2D participants had an adjusted hazard ratio of dementia of 1.56 (95% CI 1.39-2.18) compared with non-T2D participants.

    “The hazard ratio for dementia diagnosis in the new-onset T2D participants compared with the non-T2D participants was 1.56 (95%CI: 1.39-2.18).”
    • Also stated in: § Comparison of Demographic Characteristics and Risk of Dementia Between T2D and Non-T2D Participants (Analysis One) · plain
      “Cox proportional hazards regression analysis showed that the adjusted hazard ratio (HR) of dementia in T2D participants was 1.56 (95% confidence interval (CI): 1.39-2.18)when comparing with non-T2D participants (Table 1).”
    Quote in the paperSubstantial claim
  • Findingown data

    Among T2D participants taking solely one type of antidiabetic medication, TZD users had a substantially higher rate of dementia (relative rate 5.31, 95% CI 1.89-14.96) compared to MET users.

    “The relative rate of dementia was 5.31 (95% CI: 1.89-14.96)for participants taking thiazolidinediones (n = 28) and 1.22 (95% CI: 0.78-1.91)for those taking sulfonylureas (n = 796) compared to those taking metformin (n = 1,033).”
    • Also stated in: § Comparison of Demographic Characteristics and Risk of Dementia Among T2D Participants Taking Different Types of Antidiabetic Medications (Analysis Two) · plain
      “TZD users had a substantially higher rate of dementia (5.31 [95% CI: 1.89-14.96])compared to the MET users.”
    Quote in the paperSubstantial claim
  • Study design and conductreasoning

    Selection bias is an important limitation of this study.

    “An important limitation is selection bias.”
    Quote in the paperNot a substantial claim
  • Study design and conductnone given

    The follow-up period was relatively short with regard to a disease with a chronic course like dementia.

    “Another important limitation was a relatively short follow-up period with regard to a disease with a chronic course like dementia.”
    Quote in the paperNot a substantial claim
  • May combine two claims. 8
  • Classified as probably not a substantial claim. 6

  • May combine two claims.

    Findingown data

    Among T2D participants taking solely one type of antidiabetic medication, the relative rate of dementia for SU users compared to MET users was 1.22 (95% CI 0.78-1.91), with no significant difference.

    “The relative rate of dementia was similar in the SU compared to MET users (1.22, 95% CI: 0.78-1.91)(Table 3).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The rate of dementia in T2D participants was 1.78 cases per 100 person-years, which was 1.26 times higher than in non-T2D participants.

    “The rate of dementia in T2D participants was 1.78 cases per 100 person-years (n = 147) which was 1.26 times higher than in non-T2D participants (n = 4,650).”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data

    Potential confounders were identified and examined initially in univariate models, and significant covariates included in the final model were age, sex, hypertension, hyperlipidemia, and cerebrovascular disease.

    “Significant covariates included in the final model were age, sex, hypertension, hyperlipidemia, and cerebrovascular disease.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    The analysis focused on participants taking only one type of antidiabetic medication so that the comparison among the three types could be clearer without interference from other types.

    “The analysis focused on participants taking only one type of antidiabetic medication so that the comparison among the three types of antidiabetic medication could be clearer without interference from influences of other types of antidiabetic medication.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    A recent study suggests metformin and sulfonylureas may decrease the risk of dementia.

    “a recent study suggests MET and sulfonylureas (SUs) may decrease the risk of dementia (15)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [15] 2011, Incidence of dementia is increased in type 2 diabetes and reduced by the use of sulfonylureas and metformin, J Alzheimers Dis source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    One study in cellular models raises the concern that metformin might increase the risk of late onset Alzheimer's disease by increasing the generation of β-amyloid peptides through upregulation of β-secretase.

    “One study in cellular models raises the concern that MET might increase the risk of late onset AD because it shows that MET increases the generation of β-amyloid peptides through upregulation of β-secretase (14)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [14] 2009, Antidiabetic drug metformin (GlucophageR) increases biogenesis of Alzheimer's amyloid peptides via up-regulating BACE1 transcription, Proc Natl Acad Sci USA source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Metformin reduces insulin levels, inflammation and thrombosis, and the risk of metabolic syndrome, and improves insulin sensitivity, suggesting its potential to exert protective effects on cognitive function.

    “Studies showing that metformin (MET) not only reduces insulin levels (11), inflammation and thrombosis (12), and the risk of the metabolic syndrome (13), but also improves insulin sensitivity (11) suggests its potential to exert protective effects on cognitive function.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [11] 2005, Role of insulin secretion and sensitivity in the evolution of type 2 diabetes in the diabetes prevention program: effects of lifestyle intervention and metformin, Diabetes source

      Not checked in the free run.

    • cites [12] 2005, Intensive lifestyle intervention or metformin on inflammation and coagulation in participants with impaired glucose tolerance, Diabetes source

      Not checked in the free run.

    • cites [13] 2005, The effect of metformin and intensive lifestyle intervention on the metabolic syndrome: the Diabetes Prevention Program randomized trial, Ann Intern Med source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Many studies have shown that patients tend to develop dementia in a relatively shorter period of time relative to patients without diabetes.

    “However, many studies have shown that patients tend to develop dementia in a relatively shorter period of time relative to patients without diabetes.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [4] 2006, Risk of dementia in diabetes mellitus: a systematic review, Lancet Neurol source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    One trial found rosiglitazone was effective in memory and attention, but another randomized placebo controlled trial and a Phase III trial failed to demonstrate its effectiveness in cognitive function.

    “One trial found rosiglitazone was effective in memory and attention (18) but another randomized placebo controlled trial and a Phase III trial failed to demonstrate its effectiveness in cognitive function (19,20).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [18] 2005, Preserved cognition in patients with early Alzheimer disease and amnestic mild cognitive impairment during treatment with rosiglitazone: a preliminary study, Am J Geriatr Psychiatry

      Not checked in the free run.

    • cites [19] 2006, Efficacy of rosiglitazone in a genetically defined population with mild-to-moderate Alzheimer's disease, Pharmacogenomics J source

      Not checked in the free run.

    • cites [20] 2010, A multi-center randomized proof-of-concept clinical trial applying [(1)F]FDG-PET for evaluation of metabolic therapy with rosiglitazone XR in mild to moderate Alzheimer's disease, J Alzheimers Dis

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Thiazolidinediones have been consistently shown to have a higher risk of congestive heart failure, myocardial infarction, and other major cardiovascular adverse effects that all contribute to the development of dementia.

    “TZDs have been consistently shown to have a higher risk of congestive heart failure, myocardial infarction, and other major cardiovascular adverse effects that all contribute to the development of dementia (16,17).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [16] 2005, Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial, Lancet source

      Not checked in the free run.

    • cites [17] 2007, Pioglitazone and risk of cardiovascular events in patients with type 2 diabetes mellitus: a meta-analysis of randomized trials, JAMA source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Trials have failed to demonstrate thiazolidinedione effectiveness in dementia patients' cognitive function, probably due to adverse cardiovascular effects outweighing potential benefits.

    “Trials have failed to demonstrate its effectiveness in dementia patients' cognitive function (19,20), probably due to its adverse cardiovascular effects outweighing potential benefits.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [19] 2006, Efficacy of rosiglitazone in a genetically defined population with mild-to-moderate Alzheimer's disease, Pharmacogenomics J source

      Not checked in the free run.

    • cites [20] 2010, A multi-center randomized proof-of-concept clinical trial applying [(1)F]FDG-PET for evaluation of metabolic therapy with rosiglitazone XR in mild to moderate Alzheimer's disease, J Alzheimers Dis

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationown data

    These findings add to the growing evidence of a link between T2D, antidiabetic medication, and dementia.

    “These findings add to the growing evidence of a link between T2D, antidiabetic medication, and dementia.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    The putative mechanisms of metformin's positive effects include reducing insulin levels and improving insulin sensitivity, reducing inflammation, and reducing the risk of metabolic syndrome.

    “The putative mechanisms of MET's positive effects include (i) reducing insulin levels and improving insulin sensitivity (11), (ii) reducing inflammation (11), and (iii) reducing the risk of metabolic syndrome (25).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [11] 2005, Role of insulin secretion and sensitivity in the evolution of type 2 diabetes in the diabetes prevention program: effects of lifestyle intervention and metformin, Diabetes source

      Not checked in the free run.

    • cites [25] 2012, Aging biology and novel targets for drug discovery, J Gerontol A Biol Sci Med Sci

      Not checked in the free run.

13 of 13

  • The paper relies on this without support.

    Few studies have explored the association between antidiabetic medications and dementia, and their findings remain controversial.

    “Few studies have explored the association between antidiabetic medications and dementia, and their findings remain controversial.”

  • The paper relies on this without support.

    Thiazolidinediones have potent insulin-sensitizing and anti-inflammatory effects that suggest they may theoretically decrease the risk of Alzheimer's disease.

    “Although the pharmacological mechanisms of thiazolidinediones (TZDs),the potent insulin-sensitizing and anti-inflammatory effects, suggest that they may theoretically decrease the risk of AD”

  • The paper relies on this without support.

    Sulfonylureas act by increasing blood insulin levels and are not supposed to be advantageous for cognition on a pharmacological basis.

    “As for SUs, because they act by increasing blood insulin levels, they are not supposed to be advantageous for cognition on a pharmacological basis.”
3 of 3

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • In a large observational study of 67,731 non-diabetic participants aged ≥65, metformin reduced the risk of developing dementia.

    “In a large observational study, 67,731 participants, who had no evidence of dementia, were non-diabetic, and aged ≥ 65 were followed from January 2004 to December 2009, to observe the onset of T2DM and compare the risk of the incidence of dementia associated with the use of anti-diabetic drugs and reported that metformin reduced the risk of developing dementia (256).”

    From A Critical Review of the Evidence That Metformin Is a Putative Anti-Aging Drug That Enhances Healthspan and Extends Lifespan (2021) · cites this paper as [256]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Conflicting end dates for analysis one follow-up

    Minor

    The results state that the 67,731-person cohort was followed through December 2009, but the title of the table for that analysis says follow-up ended in June 2009.

1 of 1

Cites

Flagged first, then by citations

2 of the 26 references couldn’t be matched to a known paper.

Research works printed but not linked: 1

  1. Database NHIR, 2007, Introduction to the National Health Insurance Research Database (NHIRD)

Cited by

Sorted by citations

Related

Authors
Cheng, C., Lin, C.-H., Tsai, Y.-W., Tsai, C.-J., Chou, P.-H., & Lan, T.-H.
Affiliations
  • University of Cambridge (GB)
  • Taichung Veterans General Hospital (TW)
  • National Health Research Institutes (TW)
  • Changhua Christian Hospital (TW)
Corresponding author
Lan, Tsuo-Hung
ORCID
Venue
The Journals of Gerontology: Series A
Publisher
Oxford University Press (OUP)
Published
2014-06-04
Volume
69
Issue
10
Pages
1299-1305
ISSN
  • 1079-5006 (ISSN-L)
  • 1758-535X
DOI
10.1093/gerona/glu073
OpenAlex
W2097594110
Semantic Scholar
276b39aa67f3b7693342d8ac89071206ec0bd7e8
Cited by
167
References
26
Citations by year
2026: 5 · 2025: 10 · 2024: 4 · 2023: 18 · 2022: 20 · 2021: 21 · 2020: 24 · 2019: 23 · 2018: 15 · 2017: 8 · 2016: 9 · 2015: 5 · 2014: 3
FWCI
6.71
Type
journal article
Language
en
Open access
closed
Locations
  • The Journals of Gerontology Series A (publishedVersion)
  • PubMed (publishedVersion)
Funders
National Health Research Institutes
Topics
  • Health Sciences › Medicine › Endocrinology, Diabetes and Metabolism › Diabetes Treatment and Management
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Health Sciences › Medicine › Geriatrics and Gerontology › Pharmaceutical Practices and Patient Outcomes
Keywords
type 2 diabetes, dementia, antidiabetic drugs, thiazolidinediones, sulfonylureas, metformin, dementia risk, older adults, duration of drug use
MeSH terms
Aged, Dementia, Diabetes Mellitus, Type 2, Female, Humans, Hypoglycemic Agents, Male, Middle Aged, Proportional Hazards Models

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review