Skip to main content

Table 2 Summary of study findings

From: Efficacy and safety of metformin in the management of type 2 diabetes mellitus in older adults: a systematic review for the development of recommendations to reduce potentially inappropriate prescribing

Authors and publication year

Outcomes

Metformin cases/na (%)

Comparator cases/na (%)

Unadjusted Risk ratiob

(95% CI)

Reported Statistical comparisonc

Result favours

Tier 1 outcomes (hypoglycaemia and adverse events): comparisons against other non-specific treatments

 Cryer 2005 [31]

 

Metformin

Usual care

   

 Subgroup ≥ 65 years

Any SAE

371/2515 (14.8)

93/569 (16.3)

0.90 (0.73, 1.11)

NR

M

Tier 1 outcomes (hypoglycaemia and adverse events): comparisons against other specific treatments

 Janka 2007 [35]

 

Insulin glargine + OAD (glimepiride and metformin)

Premixed insulin

   

(n = 63 patients)

Confirmed + unconfirmed hypoglycaemia

(n = 67 patients)

    

5.6 events per p-yr.

11.4 events per p-yr

0.49 (0.41, 0.59)

p = 0.01

M

Confirmed hypoglycaemia

3.7 events per p-yr

9.1 events per p-yr

 

p = 0.008

M

Confirmed symptomatic hypoglycaemia

2.2 events per p-yr

5.0 events per p-yr

0.40 (0.33, 0.50)

p = 0.06

M

Confirmed nocturnal hypoglycaemia

0.4 events per p-yr

0.7 events per p-yr

0.44 (0.33, 0.59)

p = 0.26

M

Severe hypoglycaemia

0.0 events per p-yr

0.1 events per p-yr

 

p = 0.21

M

One or more treatment-emergent AEs

32/67 (47.8)

27/63 (42.9)

0.55 (0.27, 1.12)

  

0.14 (0.01, 2.61)

NR

C

1.11 (0.76, 1.63)

 Josephkutty et al. 1990 [36]

 

Metformin

Tolbutamide

 

NR

C

Side effects

32 side effects reported by 21 patients

15 side effects reported by 20 patients

 Schweizer et al. 2009 [43]

 

Metformin

Vildagliptin

   

AEs

83/165 (50.3)

74/167 (44.3)

1.14 (0.90, 1.43)

NR

C

SAEs

6/165 (3.6)

5/167 (3.0)

1.21 (0.38, 3.90)

NR

C

Gastrointestinal AEs

41/165 (24.8)

25/167 (15.0)

1.66 (1.06, 2.60)

NR

C

Hypoglycaemia

2/165 (1.2)

0/167 (0.0)

5.06 (0.24, 104.61)

NR

C

Tier 2 outcomes: comparisons against other non-specific treatments

 Cryer 2005 [31]

 

Metformin

Usual care

   

All-cause mortality

60/2515 (2.4)

12/569 (2.1)

1.13 (0.61, 2.09)

p = 0.878

C

 Subgroup ≥ 65 years

All-cause hospitalisations

334/2515 (13.3)

88/569 (15.5)

0.86 (0.69, 1.07)

p = 0.178

M

 Inzucchi 2005 [34]

 

Metformin

No insulin sensitizer

   

1-year mortality

246/1273 (19.3)

2014/6641 (30.3)

0.64 (0.57, 0.72)

HR = 0.92 (0.81, 1.06)

M

1-year MI readmission

210/1273 (16.5)

1247/6641 (18.8)

0.88 (0.77, 1.00)

HR = 1.02 (0.86, 1.20)

C

1-year HF readmission

435/1273 (34.2)

2859/6641 (43.1)

0.79 (0.73, 0.86)

HR = 1.06 (0.95, 1.18)

C

1-year all-cause readmission

759/1273 (59.6)

4268/6641 (64.3)

0.93 (0.88, 0.97)

HR = 1.04 (0.96, 1.13)

C

 Masoudi 2005 [39]

 

Metformin

No insulin sensitizer

   

Mortality

460/1861 (24.7)

4345/12069 (36.0)

0.69 (0.63, 0.75)

HR = 0.87 (0.78, 0.97)

M

All-cause readmission

1265/1861 (68.0)

8702/12069 (72.1)

0.94 (0.91, 0.97)

HR = 0.94 (0.89, 1.01)

M

HF readmission

1091/1861 (58.6)

7821/12069 (64.8)

0.90 (0.87, 0.94)

HR = 0.92 (0.86, 0.99)

M

Readmission for metabolic acidosis

2.3%

2.6%

P = 0.40

  

 MacDonald 2010 [38]

 

Metformin

No antidiabetic drugs

   

Mortality

155/376 (41)

733/1306 (56)

0.73 (0.65, 0.84)

OR = 0.65 (0.48, 0.87)

M

 Moore et al. 2013 [40]

 

Metformin

Not on metformin

   

(n = 35 patients)

(n = 91 patients)

Cognitive performance

NR

NR

OR = 1.75 (0.81, 3.78)

C

p = 0.158

 Roussel et al. 2010 [42]

Mortality:

Metformin

No metformin

   

Patients 65–80 years

191/3791 (5.0)

532/6768 (7.9)

0.64 (0.55, 0.75)

HR = 0.77 (0.62, 0.95), p = 0.02

M

Patients >80 years

71/598 (11.9)

220/1492 (14.7)

0.81 (0.63, 1.03)

HR = 0.92 (0.66, 1.28), p = 0.61

M

Tier 2 outcomes: comparisons against other specific treatments

 Evans 2010 [32]

 

Metformin monotherapy + combination (n = 205)

SU monotherapy

   

(n = 217)

1-year mortality

NR

NR

OR = 0.60 (0.37, 0.97)

M

Long-term mortality

NR

NR

OR = 0.67 (0.51, 0.88)

M

 Roumie 2012 [41]

 

Metformin

SU

   

(n = 64,009 patients)

(n = 54,005 patients)

 Subgroup ≥ 65 years

Hospitalization for acute MI, stroke or death

15.9 per 1,000p–yrs

24.6 per 1,000p–yrs

 

HR = 0.85 (0.78, 0.92)

M

Hospitalization for acute MI or stroke

12.9 per 1,000p–yrs

18.5 per 1,000p–yrs

 

HR = 0.88 (0.81, 0.97)

M

 Lapane 2015 [37]

Hospitalisation for hypoglycaemia

Metformin monotherapy

SU monotherapy

   

(n = 6151)

(n = 5807)

All ages

132 in 6518 p-yrs

289 in 6307 p-yrs

0.43 (0.35, 0.53)

HR = 0.42 (0.33, 0.53)

M

Age 75–84

55 in 2524 p-yrs

104 in 2455 p-yrs

0.51 (0.37, 0.71)

HR = 0.50 (0.34, 0.73)

M

Age 85+

39 in 2248 p-yrs

100 in 2167 p-yrs

0.38 (0.26, 0.54)

HR = 0.38 (0.25, 0.58)

M

Hospitalisation for fractures related to falls

All ages

180 in 6305 p-yrs

194 in 6174 p-yrs

0.94 (0.76, 1.15)

HR = 0.88 (0.69, 1.12)

M

Age 75–84

70 in 2478 p-yrs

86 in 2375 p-yrs

0.78 (0.57, 1.07)

HR = 0.73 (0.50, 1.05)

M

Age 85+

74 in 2142 p-yrs

65 in 2114 p-yrs

1.12 (0.81, 1.57)

HR = 1.05 (0.68, 1.59)

C

Falls

All ages

1844 in 4546 p-yrs

1864 in 4560 p-yrs

0.99 (0.93, 1.06)

HR = 1.02 (0.94, 1.11)

C

Age 75–84

703 in 1785 p-yrs

756 in 1693 p-yrs

0.88 (0.80, 0.98)

HR = 0.90 (0.79, 1.02)

M

Age 85+

697 in 1519 p-yrs

691 in 1547 p-yrs

1.03 (0.92, 1.14)

HR = 0.98 (0.86, 1.12)

M

 Hung 2013 [33]

 

Metformin (n = 109)

Glyburide (n = 72)

   

47 in 181 p-yrs

 Subgroup ≥ 71 years

Non-fatal CVD

30 in 414 p-yrs

Glimepiride (n = 50)

0.28 (0.18,0.44)

HR = 0.30 (0.18, 0.48)

M

Non-fatal CVD

30 in 414 p-yrs

18 in 167 p-yrs

0.67 (0.38, 1.21)

NR

M

 Inzucchi 2005 [34]

 

Metformin

Thiazolidinedione

   

Mortality

246/1273 (19.3)

237/819 (28.9)

0.67 (0.57, 0.78)

NR

M

MI readmission

210/1273 (16.5)

154/819 (18.8)

0.88 (0.73, 1.06)

NR

M

HF readmission

435/1273 (34.2)

402/819 (49.1)

0.70 (0.63, 0.77)

NR

M

All-cause readmission

759/1273 (59.6)

555/819 (67.8)

0.88 (0.82, 0.94)

NR

M

 Solomon 2009 [44]

 

Metformin

Thiazolidinediones

   

Fractures

110/4235 (2.6)

74/2347 (3.2)

SU

0.82 (0.62,1.10)

RR = 0.76 (0.56, 1.02)

M

Fractures

110/4235 (2.6)

480/13709 (3.5)

0.74 (0.60, 0.91)

NR

M

 Tzoulaki 2009 [45]

 

Metformin monotherapy

1st generation SU

monotherapy

   

>1.6mil intervals d

 Subgroup ≥ 65 years

MI

NR

NR

 

HR = 0.79 (0.65, 0.96)

M

CHF

NR

NR

 

HR = 0.76 (0.68, 0.85)

M

All-cause mortality

NR

NR

 

HR = 0.72 (0.67, 0.79)

M

  

2nd generation SU

monotherapy

   

MI

NR

NR

 

HR = 0.82 (0.74, 0.91)

M

CHF

NR

NR

 

HR = 0.85 (0.79, 0.91)

M

All-cause mortality

NR

NR

 

HR = 0.74 (0.70, 0.78)

M

  

Rosiglitazone

monotherapy

   

MI

NR

NR

 

HR = 0.85 (0.54, 1.33)

M

CHF

NR

NR

 

HR = 0.93 (0.65, 1.33)

M

All-cause mortality

NR

NR

 

HR = 0.98 (0.76, 1.27)

M

  

Rosiglitazone

combination

   

MI

NR

NR

 

HR = 0.81 (0.63, 1.06)

M

CHF

NR

NR

 

HR = 0.76 (0.61, 0.93)

M

All-cause mortality

NR

NR

 

HR = 1.10 (0.94, 1.28)

C

  

Pioglitazone alone and combined

   

MI

NR

NR

 

HR = 1.23 (0.74, 2.08)

C

CHF

NR

NR

 

HR = 0.90 (0.64, 1.26)

M

All-cause mortality

NR

NR

 

HR = 1.54 (1.15, 2.04)

C

  

Other drugs and combinations

   

MI

NR

NR

 

HR = 0.87 (0.77, 0.98)

M

CHF

NR

NR

 

HR = 0.93 (0.85, 1.01)

M

All-cause mortality

NR

NR

 

HR = 0.74 (0.70, 0.78)

M

  1. AEs adverse events; SAEs serious adverse events; MI Myocardial Infarction; CVD cardiovascular disease; CHF congestive heart failure; P-YRs patient-years; M Metformin; SU Sulfonylureas; C Comparator; CI confidence interval; HR hazard ratio; OR Odds ratio; RR risk ratio; NR Not Reported
  2. anumber of patients with the outcome/total patients unless stated otherwise, unreported counts/rates were derived from available data where possible
  3. bCalculated risk ratio unadjusted for covariates, zero cell adjustment applied where relevant
  4. cBased on reported comparison adjusted for covariates, or if not reported the unadjusted risk ratio
  5. dtotal treatment intervals across all treatments (metformin plus comparators) was >1.6 million, patients could have multiple intervals on different treatments