SynthesisScientific reports2015

Pharmacologic Therapy of Diabetes and Overall Cancer Risk and Mortality: A Meta-Analysis of 265 Studies.

Lang Wu, Jingjing Zhu, Larry J Prokop, Mohammad Hassan Murad

Open access · goldFull text readMeta-AnalysisSystematic Review
In one paragraph

Synthesis in Scientific reports, 2015. The graph read 15 numbers from its abstract, feeding 10 cells of the map, but none could be read as for or against, so it casts no vote. It also reports 5 associations that do not count as treatment evidence, such as RR 0.86 (0.83 to 0.90) for adverse events & safety. Cited by 76 papers, 10 of them syntheses that pooled it.

15numbers the graph read from it
0cells of the map it votes in
76citing papers in PubMed, 10 pooled it
8.7field-weighted citation impact, top 2% of its field
1 · What the graph read from it

What it found

Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.

Read, but not usablea number the graph found but could not read as for or against

Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 1.120.61 to 2.06
For GLP-1 agonist use, meta-analysis of 16 studies demonstrated no significant association with cancer risk (RR = 1.12, 95% CI 0.61-2.06) (Table 1h).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 0.900.49 to 1.65
Dapagliflozin was not associated with the risk of cancer (7 studies, RR = 0.90, 95% CI 0.49-1.65) (Table 1i).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 1.080.99 to 1.18
On the other hand, we did not detect significant association between sulfonylureas use and cancer mortality (12 studies, RR = 1.08, 95% CI 0.99-1.18) (Table 2c).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds 2 cells of the map
RR 0.860.83 to 0.90
Meta-analysis demonstrated that the use of metformin or thiazolidinediones was associated with a lower risk of cancer incidence (RR = 0.86, 95% CI 0.83-0.90, I(2) = 88.61%; RR = 0.93, 95% CI 0.91-0.96, I(2) = 0.00% respectively).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds 2 cells of the map
RR 0.930.91 to 0.96
Meta-analysis demonstrated that the use of metformin or thiazolidinediones was associated with a lower risk of cancer incidence (RR = 0.86, 95% CI 0.83-0.90, I(2) = 88.61%; RR = 0.93, 95% CI 0.91-0.96, I(2) = 0.00% respectively).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 0.920.82 to 1.04
Meta-analysis of 62 studies showed that DPP-4 inhibitors were not associated with the risk of cancer (RR = 0.92, 95% CI 0.82-1.04) (Table 1e).
Glycemic controlan association or prognostic statement, not a treatment comparison · t2dfeeds 3 cells of the map
RR 1.211.08 to 1.36
On the other hand, insulin, sulfonylureas and alpha glucosidase inhibitor use was associated with an increased risk of cancer incidence (RR = 1.21, 95% CI 1.08-1.36, I(2) = 96.31%; RR = 1.20, 95% CI 1.13-1.27, I(2) = 95.02%; RR = 1.10, 95% CI 1.05-1.15, I(2) = 0.00% respectively).
Glycemic controlan association or prognostic statement, not a treatment comparison · t2dfeeds 3 cells of the map
RR 1.201.13 to 1.27
On the other hand, insulin, sulfonylureas and alpha glucosidase inhibitor use was associated with an increased risk of cancer incidence (RR = 1.21, 95% CI 1.08-1.36, I(2) = 96.31%; RR = 1.20, 95% CI 1.13-1.27, I(2) = 95.02%; RR = 1.10, 95% CI 1.05-1.15, I(2) = 0.00% respectively).
Glycemic controlan association or prognostic statement, not a treatment comparison · t2dfeeds 3 cells of the map
RR 1.101.05 to 1.15
On the other hand, insulin, sulfonylureas and alpha glucosidase inhibitor use was associated with an increased risk of cancer incidence (RR = 1.21, 95% CI 1.08-1.36, I(2) = 96.31%; RR = 1.20, 95% CI 1.13-1.27, I(2) = 95.02%; RR = 1.10, 95% CI 1.05-1.15, I(2) = 0.00% respectively).
Glycemic controlan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 1.211.08 to 1.36
Insulin use was significantly associated with an increased risk of cancer (73 studies; RR = 1.21, 95% CI 1.08-1.36) (Table 1a).
Glycemic controlan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 1.190.80 to 1.77
With regard to cancer mortality, we did not detect significant association with insulin use (12 studies; RR = 1.19, 95% CI 0.80-1.77) (Table 2a).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 0.860.83 to 0.90
Meta-analysis of 66 and 12 studies respectively, demonstrated significant association between the use of metformin and the risk of cancer (RR = 0.86, 95% CI 0.83-0.90) and cancer mortality (RR = 0.70, 95% CI 0.53-0.94) (Table 1b and 2b).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 1.201.13 to 1.27
Use of sulfonylureas was associated with an increased risk of cancer (72 studies, RR = 1.20, 95% CI 1.13-1.27) (Table 1c).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds 2 cells of the map
RR 0.930.91 to 0.96
Similar to the case of metformin, TZD use was associated with a decreased risk of cancer (119 studies, RR = 0.93, 95% CI 0.91-0.96) (Table 1d).
Adverse events & safetyan association or prognostic statement, not a treatment comparison · t2dfeeds one cell of the map
RR 1.101.05 to 1.15
Alpha glucosidase inhibitor use was associated with increased risk of cancer (13 studies, RR =  1.10, 95% CI 1.05-1.15) (Table 2f).

clause the extractor read what became the number

2 · Its place on the map

Where it lands on the map

Rows are treatments, columns are outcomes. The coloured squares are the cells this paper feeds, coloured by the vote it casts there. Click one to jump to what this paper adds to it.

supports the treatmentfavours the comparatorno clear differenceread, but no usable result
3 · What it changes

What it adds to each cell

For every cell the paper feeds: the belief in the claim with and without this paper, and this paper's estimate drawn against every other readable study in the cell. The ringed dot is this paper.

DPP-4 inhibitors×adverse events & safety

No readable resultOpen on the map →What to test next →

34 readable studies in this cell: 12 favour the treatment, 17 find no difference, 5 favour the comparator.

Belief with this paper
0.00contested · 0 families support, 19 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the comparatorfavours the treatment →
0 · no effect
NCT040178321,441 enrolled · 2019
Δ -0.20-0.30 to -0.10
NCT020991101,233 enrolled · 2014
Δ -3.20-11.7 to 5.50
NCT019301881,231 enrolled · 2013
Δ -1.06-1.21 to -0.91
NCT007344741,202 enrolled · 2008
Δ -0.71-0.87 to -0.55
NCT022730501,136 enrolled · 2014
Δ 0.00-2.30 to 2.30
NCT004499301,050 enrolled · 2007
Δ -7.30-10.6 to -4.20
NCT008389031,049 enrolled · 2009
Δ -0.35-0.53 to -0.17
NCT01023581784 enrolled · 2009
Δ -0.57-0.87 to -0.27
NCT01682759751 enrolled · 2012
Δ -6.90-13.9 to 0.10
NCT02738879746 enrolled · 2016
Δ -2.10-9.10 to 5.00
NCT01217073685 enrolled · 2010
Δ 6.20-6.20 to 18.4
NCT01890122647 enrolled · 2013
Δ -0.49-0.70 to -0.28

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

GLP-1 receptor agonists×adverse events & safety

No readable resultOpen on the map →What to test next →

40 readable studies in this cell: 47 favour the treatment, 14 find no difference, 2 favour the comparator.

Belief with this paper
0.02contested · 1 family supports, 41 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the treatmentfavours the comparator →
0 · no effect
NCT017204463,297 enrolled · 2013
Δ -0.66-0.80 to -0.52
NCT036893742,274 enrolled · 2018
Δ -0.29-0.38 to -0.20
NCT013360231,663 enrolled · 2011
Treatment contrast -0.64-0.75 to -0.53
NCT040178321,441 enrolled · 2019
Δ -0.20-0.30 to -0.10
NCT018365231,398 enrolled · 2013
Δ -0.20-0.32 to -0.07
NCT019301881,231 enrolled · 2013
Δ -1.06-1.21 to -0.91
NCT007344741,202 enrolled · 2008
Δ -0.71-0.87 to -0.55
NCT008389031,049 enrolled · 2009
Δ -0.91-1.16 to -0.65
NCT009606611,036 enrolled · 2009
Δ -0.04-0.18 to 0.11
NCT050350821,018 enrolled · 2021
Δ -0.24-0.44 to -0.04
NCT01064687978 enrolled · 2010
Δ -1.05-1.22 to -0.88
NCT01117350978 enrolled · 2010
Δ 2.54-3.88 to 8.93

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Insulin×glycemic control

No readable resultOpen on the map →What to test next →

40 readable studies in this cell: 15 favour the treatment, 14 find no difference, 14 favour the comparator.

Belief with this paper
0.50contested · 9 families support, 6 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the comparatorfavours the treatment →
0 · no effect
NCT036893742,274 enrolled · 2018
Δ -0.29-0.38 to -0.20
NCT037306622,002 enrolled · 2018
Δ -0.99-1.13 to -0.86
NCT013360231,663 enrolled · 2011
Treatment contrast -0.64-0.75 to -0.53
NCT038829701,444 enrolled · 2019
Δ -0.86-1.00 to -0.72
NCT045379231,428 enrolled · 2020
Δ -1.10-1.24 to -0.97
NCT032680051,264 enrolled · 2017
Δ -0.04-0.11 to 0.03
NCT020581471,170 enrolled · 2014
Δ -0.78-0.90 to -0.67
NCT021289321,089 enrolled · 2014
Δ -0.81-0.96 to -0.67
NCT009606611,036 enrolled · 2009
Δ -0.04-0.18 to 0.11
NCT00856986987 enrolled · 2009
Δ -0.52-0.68 to -0.36
NCT01117350978 enrolled · 2010
Δ 2.54-3.88 to 8.93
NCT03214380933 enrolled · 2017
Δ 0.06-0.05 to 0.16

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Metformin×adverse events & safety

No readable resultOpen on the map →What to test next →

22 readable studies in this cell: 9 favour the treatment, 8 find no difference, 5 favour the comparator.

Belief with this paper
0.27contested · 4 families support, 11 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the treatmentfavours the comparator →
0 · no effect
NCT017190031,413 enrolled · 2012
Adjusted mean -0.72-0.95 to -0.48
NCT018093271,186 enrolled · 2013
Δ -0.40-0.59 to -0.21
NCT022730501,136 enrolled · 2014
Δ -8.90-13.3 to -4.50
NCT01126580807 enrolled · 2010
Δ -0.22-0.36 to -0.08
NCT01023581784 enrolled · 2009
Δ -0.67-0.96 to -0.37
NCT00386100688 enrolled · 2006
Δ 1.47-14.8 to 20.8
NCT01217073685 enrolled · 2010
Δ 1.00-9.80 to 13.1
NCT01890122647 enrolled · 2013
Δ -0.68-0.89 to -0.47
NCT00727857600 enrolled · 2007
Δ 0.860.51 to 1.22
NCT01958671461 enrolled · 2013
Δ -2.60-13.1 to 8.10
NCT00328172302 enrolled · 2006
Δ -0.85-1.10 to -0.59
NCT01485614200 enrolled · 2012
Δ 2.40-10.0 to 14.9

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Other glucose-lowering×glycemic control

No readable resultOpen on the map →What to test next →

8 readable studies in this cell: 5 favour the treatment, 0 find no difference, 3 favour the comparator.

Belief with this paper
0.50contested · 3 families support, 1 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the treatmentfavours the comparator →
0 · no effect
NCT017093055,570 enrolled · 2012
Δ 0.190.02 to 0.36
NCT050350821,018 enrolled · 2021
Δ -0.24-0.44 to -0.04
placebo-subtracted decrease -0.44

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Other glucose-lowering×adverse events & safety

No readable resultOpen on the map →What to test next →

3 readable studies in this cell: 0 favour the treatment, 1 find no difference, 2 favour the comparator.

Belief with this paper
0.50one trial · 1 family supports, 0 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the comparatorfavours the treatment →
0 · no effect
NCT017093055,570 enrolled · 2012
Estimate 0.40-0.30 to 1.30
NCT050350821,018 enrolled · 2021
Δ -0.24-0.44 to -0.04

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

SGLT2 inhibitors×adverse events & safety

No readable resultOpen on the map →What to test next →

38 readable studies in this cell: 18 favour the treatment, 19 find no difference, 1 favour the comparator.

Belief with this paper
0.00contested · 0 families support, 26 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the treatmentfavours the comparator →
0 · no effect
NCT011374742,996 enrolled · 2010
Δ -0.46-0.59 to -0.33
NCT017190031,413 enrolled · 2012
Adjusted mean -0.33-0.56 to -0.10
NCT020991101,233 enrolled · 2014
Δ 1.40-7.40 to 10.1
NCT018093271,186 enrolled · 2013
Δ -0.46-0.66 to -0.27
NCT01649297983 enrolled · 2012
Δ -0.61-0.79 to -0.44
NCT02580591977 enrolled · 2015
Δ -0.28-0.46 to -0.11
NCT02414958730 enrolled · 2015
Δ -0.54-0.66 to -0.41
NCT01381900678 enrolled · 2011
Δ -0.51-0.64 to -0.37
NCT02033889621 enrolled · 2013
Δ 1.50-2.10 to 5.40
NCT02532855614 enrolled · 2015
Δ -2.80-10.7 to 5.10
NCT02630706506 enrolled · 2015
Δ -6.00-16.5 to 4.60
NCT02036515464 enrolled · 2014
Δ -5.70-16.5 to 5.20

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Sulfonylureas & glinides×adverse events & safety

No readable resultOpen on the map →What to test next →

8 readable studies in this cell: 1 favour the treatment, 6 find no difference, 1 favour the comparator.

Belief with this paper
0.00contested · 0 families support, 1 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the comparatorfavours the treatment →
0 · no effect
NCT017093055,570 enrolled · 2012
Estimate 0.40-0.30 to 1.30
NCT008389031,049 enrolled · 2009
Δ -0.27-0.45 to -0.09
NCT01682759751 enrolled · 2012
Δ -6.90-13.9 to 0.10
NCT02033889621 enrolled · 2013
Δ 1.50-2.10 to 5.40
NCT00770653305 enrolled · 2007
Δ 28.1-95.6 to 152
NCT0062028249 enrolled · 2008
Least squares mean 0.71-6.90 to 8.32
NCT0297347745 enrolled · 2017
Coefficient -0.00-0.16 to 0.15

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Sulfonylureas & glinides×glycemic control

No readable resultOpen on the map →What to test next →

31 readable studies in this cell: 7 favour the treatment, 13 find no difference, 11 favour the comparator.

Belief with this paper
0.83established · 5 families support, 1 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the comparatorfavours the treatment →
0 · no effect
NCT017093055,570 enrolled · 2012
Δ 0.190.02 to 0.36
NCT009688121,452 enrolled · 2009
Δ -0.01-0.11 to 0.09
NCT006609071,217 enrolled · 2008
Δ 0.00-0.11 to 0.11
NCT003184611,091 enrolled · 2006
Δ -0.02-0.19 to 0.15
NCT008389031,049 enrolled · 2009
Δ -0.27-0.45 to -0.09
Δ 0.640.33 to 0.95
NCT03332771954 enrolled · 2017
Δ 0.12-0.12 to 0.36
NCT02471404939 enrolled · 2015
Δ 0.160.03 to 0.30
NCT00614120929 enrolled · 2008
Δ -0.06-0.23 to 0.11
NCT00575588891 enrolled · 2007
Δ 0.06-0.05 to 0.16
NCT01682759751 enrolled · 2012
Δ 0.180.06 to 0.30
NCT00294723746 enrolled · 2006
Δ -0.62-0.83 to -0.42

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

Thiazolidinediones×adverse events & safety

No readable resultOpen on the map →What to test next →

6 readable studies in this cell: 3 favour the treatment, 2 find no difference, 1 favour the comparator.

Belief with this paper
0.00contested · 0 families support, 2 contradict · against placebo
Without itNot a counted family in this claim, so removing it changes nothing.
← favours the comparatorfavours the treatment →
0 · no effect
NCT00839527685 enrolled · 2009
Δ 0.250.10 to 0.40
NCT00727857600 enrolled · 2007
Δ 0.840.50 to 1.18
NCT00770653305 enrolled · 2007
Δ 28.1-95.6 to 152
NCT0158944577 enrolled · 2008
Δ -0.64-4.40 to 2.40

This paper's own estimate is on a different scale from the rest of the cell, so it is not drawn here.

4 · The registry

The trial behind it

Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.

Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.

5 · Its place in the literature

Who cites it

76 citing papers in PubMed, 10 syntheses or guidelines pooled it, 143 citations in OpenAlex.

  1. Pooled it
  2. Pooled it
  3. American Association of Clinical Endocrinology Clinical Practice Guideline: Developing a Diabetes Mellitus Comprehensive Care Plan-2022 Update.Endocrine practice : official journal of the American College of Endocrinology and the American Association of Clinical Endocrinologists · 2022
    Guideline
  4. Pooled it
  5. Pooled it
  6. Pooled it
  7. Pooled it
  8. Efficacy of Ligustrazine Injection as Adjunctive Therapy for Angina Pectoris: A Systematic Review and Meta-Analysis.Medical science monitor : international medical journal of experimental and clinical research · 2015
    Pooled it
  9. Cyclooxygenase-2-1195G>A Polymorphism and Head and Neck Squamous Cell Carcinoma Susceptibility: A Meta-Analysis of 1564 Cases and 2346 Controls.Medical science monitor : international medical journal of experimental and clinical research · 2015
    Pooled it
  10. Pooled it
  11. Review
  12. Review
  13. Article
  14. Review
  15. Article
  16. Pan-cancer analysis and oncogenic implications ofJournal of cell communication and signaling · 2025
    Article
  17. Article
  18. Article
  19. Associations between Diabetes Mellitus and Selected Cancers.International journal of molecular sciences · 2024
    Review
  20. Review

16 more citing papers are in PubMed but not listed here.

6 · The record

Corrections and comments

PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.

7 · Who and what money

Authors and funding

4 authors at 3 institutions in 1 country.

Lang WuCenter for Clinical and Translational Science, Mayo Clinic, Rochester, Minnesota.
Jingjing ZhuProgram of Quantitative Methods in Education, University of Minnesota, Minneapolis, Minnesota.
Larry J ProkopMayo Clinic Libraries, Mayo Clinic, Rochester, Minnesota.
Mohammad Hassan Murad1] Division of Preventive Medicine, Mayo Clinic, Rochester, Minnesota [2] Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, Minnesota.
Mayo Clinic · USUniversity of Minnesota · USWinnMed · US

Funding

No grant is acknowledged in the PubMed record.

8 · The paper itself

Abstract

The marked sentences are the ones the graph read a number from.

Different anti-diabetic medications (ADMs) may modify cancer risk and mortality in patients with diabetes. We conducted a systematic review and meta-analysis to estimate the magnitude of association and quality of supporting evidence for each ADM. A total of 265 studies (44 cohort studies, 39 case-control studies, and 182 randomized controlled trials (RCT)) were identified, involving approximately 7.6 million and 137,540 patients with diabetes for observational studies and RCTs, respectively. The risk of bias overall was moderate. Meta-analysis demonstrated that the use of metformin or thiazolidinediones was associated with a lower risk of cancer incidence (RR = 0.86, 95% CI 0.83-0.90, I(2) = 88.61%; RR = 0.93, 95% CI 0.91-0.96, I(2) = 0.00% respectively). On the other hand, insulin, sulfonylureas and alpha glucosidase inhibitor use was associated with an increased risk of cancer incidence (RR = 1.21, 95% CI 1.08-1.36, I(2) = 96.31%; RR = 1.20, 95% CI 1.13-1.27, I(2) = 95.02%; RR = 1.10, 95% CI 1.05-1.15, I(2) = 0.00% respectively). Use of other types of ADMs was not significantly associated with cancer risk. This study indicates that some ADMs may modify the risk of cancer in individuals with diabetes. Knowledge of this risk may affect the choice of ADM in individuals concerned about cancer or at increased risk for cancer.

Indexed as

Diabetes Mellitus, Type 2Glycoside Hydrolase InhibitorsHumansHypoglycemic AgentsInsulinMetforminNeoplasmsOdds RatioRandomized Controlled Trials as TopicRiskSulfonylurea CompoundsThiazolidinedionesGlycoside Hydrolase InhibitorsHypoglycemic AgentsInsulinMetforminSulfonylurea CompoundsThiazolidinediones

Identifiers

PMID26076034
PMCPMC4467243
OpenAlexW2262375108

What Socratic holds

Textfull text, public
LicenceCC BY
Read underepoch 390

Registered trials

None linked

Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the Socratic graph.