Table 1 Summary of study characteristics

Cryer 2005 [31]

Randomised, open label, parallel-group, multicentre, clinical trial

To evaluate the risk of lactic acidosis or other SAEs with metformin, under usual care conditions.

P: 8732

P ≥ 65 years: 3084

P using metformin (metformin only or in combination with SU or non-SU oral agent): 2515

P under usual care (SU, TZD, insulin, or any other non-metformin monotherapy or combination therapy): 569

12 months

Incidence of SAEs^c, hospitalization, and death.

Evans 2010 [32]

Population-based prospective cohort study

To examine the efficacy of metformin and SU in patients with DM2 and CHF.

P: 422

P ≥ 65 years: 365

P using SU: 191

P using any metformin (alone/combination): 174

1 year and at end of follow-up (death, loss to follow-up, or end of 10 year study window)

All-cause mortality

Hung 2013 [33]

Population-based retrospective cohort study

To assess the risk of non-fatal cardiovascular events among patients with DM2 who are taking metformin monotherapy, glimepiride or glyburide.

P: 1159

P ≥ 71 years: 231

P using glyburide: 72

P using glimepiride: 50

P using metformin: 109

3 months

Incidence of non-fatal cardiovascular events including coronary artery disease, peripheral artery disease, stroke and heart failure.

Inzucchi 2005 [34]

Retrospective cohort study

To determine the impact of insulin sensitizers on outcomes in diabetic patients after hospitalization with AMI.

P: 8872

P ≥ 65 years: 8872

P using no insulin sensitizer: 6641

P using metformin: 1273

P using TZD: 819

P using metformin + TZD: 139

1 year

Time from hospital discharge to death from any cause censored at 1 year of discharge.

Time to first readmission for MI, first readmission for heart failure, and first readmission for any cause up to 1 year after discharge.

Janka 2007 [35]

Parallel-group, open-label, randomized, multinational clinical trial

To investigate whether the safety and efficacy benefits of initiating insulin therapy with glargine and continued OADs, versus switching to premixed insulin, as previously reported, were also observed in the subset of patients aged 65 and older.

P: 130

P ≥ 65 years: 130

P using insulin glargine + glimepiride and metformin: 67

P using premixed insulin: 63

24 weeks

Hypoglycaemic events and their frequency.

Josephkutty et al. 1990 [36]

Randomised double-blind cross-over study

Efficacy, metabolic effects, and acceptability of metformin were compared with tolbutamide in 20 diabetic patients aged between 65 and 95 years.

P: 20

P ≥ 65 years: 20

P using metformin/tolbutamide: 10

P using tolbutamide/metformin: 10

3 months with each treatment

Side-effects of drug treatment.

Lapane et al. 2015 [37]

Retrospective cohort study

To evaluate the extent to which SU use was associated with fractures and falls among nursing home residents with DM2.

P: 11,958

P ≥ 65 years: 10,916

P using SU: 5128

P using biguanide: 5788

Median 683 days (range: 1–1002 days)

Severe hypoglycaemia, falls, and fractures occurring in parts of the body typically associated with falling.

MacDonald 2010 [38]

Case-control study

To examine outcomes in patients with DM2 and heart failure and to determine whether outcomes were associated with antidiabetic drug therapy.

P: 3266

P ≥ 65 years: 3102

Cases (P with DM2 and heart failure who died): 1633

Controls (P with DM2 and CHF alive): 1633

Average 11 years

All-cause mortality.

Masoudi 2005 [39]

Retrospective cohort study

To assess the relationship between the prescription of insulin-sensitizing agents (metformin and/or TZD) and death or readmission ofelderly diabetic patients initially admitted with heart failure in a cohort derived from the National Heart Care Project (NHC).

P: 5296

P ≥ 65 years: 5296

P using metformin: 1861

P using TZD: 2226

P using no insulin sensitizer: 12,069

1 year

Time from hospital discharge to death due to any cause, time to first readmission for any cause or for heart failure, proportions of patients who died or were readmitted at least once in the year after discharge, rates of readmission for the primary diagnosis of metabolic acidosis.

Moore et al. 2013 [40]

Cross sectional study

To investigate the associations of metformin, serum vitamin B12, calcium supplements, and cognitive impairment in patients with DM2.

P: 1354

P ≥ 65 years: 1164

P with DM2/impaired glucose tolerance: 126

P using metformin: 35

P no using metformin: 91

Not applicable

Cognitive performance measured with the Mini-Mental State Examination (MMSE).

Roumie 2012 [41]

Retrospective cohort study

To compare the effects of SU and metformin monotherapy on CVD outcomes (AMI and stroke) or death.

P: 253,640

P ≥ 65 years: 118,014

P using metformin: 64,009

P using SU: 54,005

12 months

Hospitalization for AMI or stroke, or death. Composite of AMI and stroke events only.

Roussel et al. 2010 [42]

Prospective, observational registry

To assess whether metformin use was associated with difference in mortality after adjustment for baseline differences and for the propensity to receive metformin among diabetics with established coronary artery disease, cerebrovascular disease, or peripheral arterial disease.

P: 19,553

P ≥ 65 years: 12,649

P using metformin: 4389

P without metformin: 8260

2 years

2-year all-cause mortality. Cardiovascular death and first-occurring event among death, MI, or stroke.

Schweizer 2009 et al. [43]

Randomized, active-controlled, parallel-group study

To compare the efficacy and tolerability of vildagliptin with metformin in elderly patients with DM2.

P: 335

P ≥ 65 years: 335

P using metformin: 166

P using vildagliptin: 169

24 weeks

AEs.

Solomon 2009 [44]

Retrospective cohort study

To determine the association between TZD use and fracture risk among older adults with DM2.

P: 20,291

P ≥ 65 years: 20,291

P using TZD: 2347

P using SU: 13,709

P using metformin: 4235

Follow-up ended at the first of any of the following events: death, loss of eligibility for Medicare or the drug benefit program, 180 days after the last dosage of oral hypoglycaemic agent, or end of follow-up.

Incidence of fracture within the cohort.

Tzoulaki 2009 [45]

Retrospective cohort study

To investigate the risk of incident MI, congestive heart failure, and all-cause mortality associated with prescription of OADs monotherapies and combinations.

P: 91,521

P ≥ 65 years: 45,760

P using FGS: 4764^b

P using SGS: 40,266 ^b

P using rosiglitazone: 4437^b

P using rosiglitazone combination: 4642^b

P using pioglitazone (alone/combination): 1879^b

P using other drugs or combinations: 21,994^b

P using metformin: 37,357^b

Mean follow-up 7.1 years

First occurrence of incident MI, CHF, and all-cause mortality.