Antidepressant use and new-onset diabetes: A systematic review and meta-analysis

Sandipan Bhattacharjee, Rituparna Bhattacharya, George A. Kelley, Usha Sambamoorthi

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

Antidepressant use has been linked to new-onset diabetes. However, the existing literature on this relationship has yielded inconsistent findings. The primary objective of this study was to systematically synthesize the literature on the relationship between antidepressant use and new-onset diabetes using meta-analysis. A systematic literature search was conducted to identify relevant studies in seven electronic databases. Two independent reviewers identified the final list of studies to be included in the meta-analysis using a priori selection criteria. Results for the primary outcome of interest, that is, odds and hazards of developing new-onset diabetes, were pooled using a random-effects model. Egger's regression test and the Trim and Fill method were utilized to detect the presence of any potential publication bias. Sensitivity analysis was conducted using the leave-one-out method as well as individual categories of antidepressant drugs. Eight studies met the inclusion criteria. Random effects models revealed that adults with any use of antidepressants were more likely to develop new-onset diabetes compared with those without any use of antidepressants [odd ratios=1.50, 95% confidence interval (CI), 1.08-2.10; hazards ratio=1.19, 95% CI, 1.08-1.32]. Sensitivity analyses revealed fair robustness; selective serotonin reuptake inhibitors and tricyclic antidepressants were more likely to be associated with the development of new-onset diabetes. Results from the Egger's regression test and Trim and Fill method revealed no evidence of publication bias. Among adults, antidepressant use was associated with higher chances of new-onset diabetes. However, because a cause-and-effect relationship cannot be established by observational studies, future randomized controlled studies are needed to confirm this association.

Original languageEnglish (US)
Pages (from-to)273-284
Number of pages12
JournalDiabetes/Metabolism Research and Reviews
Volume29
Issue number4
DOIs
StatePublished - May 2013
Externally publishedYes

Fingerprint

Antidepressive Agents
Meta-Analysis
Publication Bias
Confidence Intervals
Tricyclic Antidepressive Agents
Serotonin Uptake Inhibitors
Patient Selection
Observational Studies
Odds Ratio
Databases

Keywords

  • Antidepressants
  • Diabetes
  • Meta-analysis
  • New-onset
  • Systematic review

ASJC Scopus subject areas

  • Endocrinology
  • Endocrinology, Diabetes and Metabolism
  • Internal Medicine

Cite this

Antidepressant use and new-onset diabetes : A systematic review and meta-analysis. / Bhattacharjee, Sandipan; Bhattacharya, Rituparna; Kelley, George A.; Sambamoorthi, Usha.

In: Diabetes/Metabolism Research and Reviews, Vol. 29, No. 4, 05.2013, p. 273-284.

Research output: Contribution to journalArticle

Bhattacharjee, Sandipan ; Bhattacharya, Rituparna ; Kelley, George A. ; Sambamoorthi, Usha. / Antidepressant use and new-onset diabetes : A systematic review and meta-analysis. In: Diabetes/Metabolism Research and Reviews. 2013 ; Vol. 29, No. 4. pp. 273-284.
@article{65d7fe69d74a4415825455045d42eeb5,
title = "Antidepressant use and new-onset diabetes: A systematic review and meta-analysis",
abstract = "Antidepressant use has been linked to new-onset diabetes. However, the existing literature on this relationship has yielded inconsistent findings. The primary objective of this study was to systematically synthesize the literature on the relationship between antidepressant use and new-onset diabetes using meta-analysis. A systematic literature search was conducted to identify relevant studies in seven electronic databases. Two independent reviewers identified the final list of studies to be included in the meta-analysis using a priori selection criteria. Results for the primary outcome of interest, that is, odds and hazards of developing new-onset diabetes, were pooled using a random-effects model. Egger's regression test and the Trim and Fill method were utilized to detect the presence of any potential publication bias. Sensitivity analysis was conducted using the leave-one-out method as well as individual categories of antidepressant drugs. Eight studies met the inclusion criteria. Random effects models revealed that adults with any use of antidepressants were more likely to develop new-onset diabetes compared with those without any use of antidepressants [odd ratios=1.50, 95{\%} confidence interval (CI), 1.08-2.10; hazards ratio=1.19, 95{\%} CI, 1.08-1.32]. Sensitivity analyses revealed fair robustness; selective serotonin reuptake inhibitors and tricyclic antidepressants were more likely to be associated with the development of new-onset diabetes. Results from the Egger's regression test and Trim and Fill method revealed no evidence of publication bias. Among adults, antidepressant use was associated with higher chances of new-onset diabetes. However, because a cause-and-effect relationship cannot be established by observational studies, future randomized controlled studies are needed to confirm this association.",
keywords = "Antidepressants, Diabetes, Meta-analysis, New-onset, Systematic review",
author = "Sandipan Bhattacharjee and Rituparna Bhattacharya and Kelley, {George A.} and Usha Sambamoorthi",
year = "2013",
month = "5",
doi = "10.1002/dmrr.2393",
language = "English (US)",
volume = "29",
pages = "273--284",
journal = "Diabetes/Metabolism Research and Reviews",
issn = "1520-7552",
publisher = "John Wiley and Sons Ltd",
number = "4",

}

TY - JOUR

T1 - Antidepressant use and new-onset diabetes

T2 - A systematic review and meta-analysis

AU - Bhattacharjee, Sandipan

AU - Bhattacharya, Rituparna

AU - Kelley, George A.

AU - Sambamoorthi, Usha

PY - 2013/5

Y1 - 2013/5

N2 - Antidepressant use has been linked to new-onset diabetes. However, the existing literature on this relationship has yielded inconsistent findings. The primary objective of this study was to systematically synthesize the literature on the relationship between antidepressant use and new-onset diabetes using meta-analysis. A systematic literature search was conducted to identify relevant studies in seven electronic databases. Two independent reviewers identified the final list of studies to be included in the meta-analysis using a priori selection criteria. Results for the primary outcome of interest, that is, odds and hazards of developing new-onset diabetes, were pooled using a random-effects model. Egger's regression test and the Trim and Fill method were utilized to detect the presence of any potential publication bias. Sensitivity analysis was conducted using the leave-one-out method as well as individual categories of antidepressant drugs. Eight studies met the inclusion criteria. Random effects models revealed that adults with any use of antidepressants were more likely to develop new-onset diabetes compared with those without any use of antidepressants [odd ratios=1.50, 95% confidence interval (CI), 1.08-2.10; hazards ratio=1.19, 95% CI, 1.08-1.32]. Sensitivity analyses revealed fair robustness; selective serotonin reuptake inhibitors and tricyclic antidepressants were more likely to be associated with the development of new-onset diabetes. Results from the Egger's regression test and Trim and Fill method revealed no evidence of publication bias. Among adults, antidepressant use was associated with higher chances of new-onset diabetes. However, because a cause-and-effect relationship cannot be established by observational studies, future randomized controlled studies are needed to confirm this association.

AB - Antidepressant use has been linked to new-onset diabetes. However, the existing literature on this relationship has yielded inconsistent findings. The primary objective of this study was to systematically synthesize the literature on the relationship between antidepressant use and new-onset diabetes using meta-analysis. A systematic literature search was conducted to identify relevant studies in seven electronic databases. Two independent reviewers identified the final list of studies to be included in the meta-analysis using a priori selection criteria. Results for the primary outcome of interest, that is, odds and hazards of developing new-onset diabetes, were pooled using a random-effects model. Egger's regression test and the Trim and Fill method were utilized to detect the presence of any potential publication bias. Sensitivity analysis was conducted using the leave-one-out method as well as individual categories of antidepressant drugs. Eight studies met the inclusion criteria. Random effects models revealed that adults with any use of antidepressants were more likely to develop new-onset diabetes compared with those without any use of antidepressants [odd ratios=1.50, 95% confidence interval (CI), 1.08-2.10; hazards ratio=1.19, 95% CI, 1.08-1.32]. Sensitivity analyses revealed fair robustness; selective serotonin reuptake inhibitors and tricyclic antidepressants were more likely to be associated with the development of new-onset diabetes. Results from the Egger's regression test and Trim and Fill method revealed no evidence of publication bias. Among adults, antidepressant use was associated with higher chances of new-onset diabetes. However, because a cause-and-effect relationship cannot be established by observational studies, future randomized controlled studies are needed to confirm this association.

KW - Antidepressants

KW - Diabetes

KW - Meta-analysis

KW - New-onset

KW - Systematic review

UR - http://www.scopus.com/inward/record.url?scp=84877320578&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84877320578&partnerID=8YFLogxK

U2 - 10.1002/dmrr.2393

DO - 10.1002/dmrr.2393

M3 - Article

C2 - 23390036

AN - SCOPUS:84877320578

VL - 29

SP - 273

EP - 284

JO - Diabetes/Metabolism Research and Reviews

JF - Diabetes/Metabolism Research and Reviews

SN - 1520-7552

IS - 4

ER -