Table 1 Study sample, randomization, dropouts, and frailty criteria of all 28 articles

Alexander et al. [39]

n = 161 community-dwelling elderly (≥65 years, mean age 82 years)

Exercise group (EG) = 81 persons and Control group (CG) = 80

n = 37, EG = 21, and CG = 16

Requiring assistance in performing ≥1 mobility-related ADL(transferring, walking, bathing, and going to the toilet)

Binder et al. [14]

n = 119 (115 followed-up) community-dwelling older adults (≥78 years, mean age of 83 ± 4 years)

3:2 ratio (EG and CG, respectively); EG = 69 and CG = 50

n = 32, CG = 9, and EG = 23

Mild to moderate frailty—meeting ≥2 indicators: (1) score between 12 and 32 on the modified physical performance test (PPT); (2) difficulty or need ≥2 IADL or 1 ADL; and (3) VO2 peak between 10 and 18 ml/kg/min

Binder et al. [37]

n = 90 (85 followed-up) community-dwelling older adults (≥65 years) with proximal femur fracture (within 16 weeks from screening evaluation)

Within strata based on surgical repair procedure; CG = 44 and EG = 46

n = 22, CG = 8, and EG = 14

Modified PPT score between 12 and 28 and self-reported difficulty or requirement for assistance with ≥1 ADL

Brown et al. [15]

n = 87 (men = 37) community-dwelling independent-living (but with difficulty) older adults (≥78 years, mean age 83 ± 4 years)

EG = 48 and CG = 39

PPT Score between 18 and 31 (extreme values included)

Chin A Paw et al. [16]

n = 217 (70% women) community-dwelling elderly (≥70 years, mean age of 78.7)

Placebo-controlled intervention based on a 2 × 2 factorial design; EG (n = 55); a group receiving enriched foods (EF, n = 58); EG + EF (n = 60); or CG (n = 44)

n = 56, 26–29% in intervention groups and 16% in CG

inactivity (weekly ≥30 min of brisk walking, cycling, and gymnastics) and involuntary weight loss or a body mass index (BMI) below average

Chin A Paw et al. [17]

n = 139 community-dwelling independent-living elderly (≥70 years, mean age 78.5 ± 5.7)

see Chin A Paw et al. [16]

see Chin A Paw et al. [16]

de Jong et al. [18]

n = 143 community-dwelling independent-living elderly (≥70 years, mean age 78.6 ± 5.6)

see Chin A Paw et al. [16]

see Chin A Paw et al. [16]

Dorner et al. [27]

n = 42 long-term facility care residents(≥75 years, mean age 86.8 ± 5.8)

Within strata based on sex, age, mini mental state exam score, and strength (EG = 21 and CG = 21)

n = 12, EG = 6, and CG = 6

Impaired mobility (Tinetti score)

Ehsani et al. [19]

n = 46 older adults extracted from Binder et al. 2002

EG = 22 (83 ± 3.6 years) who had increased their aerobic capacity due to training and CG = 24 (84 ± 4.2 years)

Binder et al. [14]

Faber et al. [20]

n = 278 (79% women) residents of self-care and nursing care residences (63–98 years old, mean age 84.9 ± 6 years)

Long-term care centers were randomized to one of the two exercise programs; participants of each center were randomized to EG and CG

n = 40

Meeting ≥3 frailty indicators (≤2 for prefrailty): (1)unintentional weight loss, (2)weakness, (3)exhaustion, (4)slowness, and (5)low physical activity

Gill et al. [21]

n = 188 elderly (≥75 years, mean age 83 years) who lived at home

Within strata based on the level of physical frailty; EG = 94 and CG = 94

n = 43, EG = 33, and CG = 10

Meeting one (moderate) or two (severe) frailty indicators: (1) requiring more than 10 s to perform a gait test and (2) not being able to stand up from a seated position with their arms folded

Gill et al. [40]

see Gill et al. [21]

see Gill et al. [21]

see Gill et al. [21]

see Gill et al. [21]

Greenspan et al. [22]

n = 267 transitionally frail older women (≥70 years); residents of independent-living facilities

EG = 148 and CG = 143

n = 86, EG = 45, and CG = 41

Speechley and Tinetti (1991) criteria based on age, gait, and balance, walking activity for exercise, other physical activity for exercise, presence or absence of depression, use of sedatives, near–vision status, upper and lower extremity strength, and lower extremity disability

Greiwe et al. [23]

n = 13 older adults (>75 years)

EG = 8 and CG = 5

n = 0

Based on functional fitness tests

Jensen et al. [28]

n = 187(75% women) elderly (65–98 years; median age 84) living in care facilities

Cluster randomization of the facility care residences with EG = 89 and CG = 98

n = 35, EG = 12, and CG = 23

being most prone to falls (mobility interaction fall chart)

Latham et al. [24]

n = 243 elderly (≥65 years; mean age 79.1 ± 6.9 years)

2 × 2 stratified block randomization technique (6/block): EG = 120 and CG = 123; and Vitamin = 121, and placebo = 122

n = 21, EG = 8, and CG = 13

Based on Winograd et al. (1991) criteria, i.e., meeting ≥1 indicators: (1) dependency in ADL; (2) prolonged bed rest; (3) impaired mobility; and (4) recent fall

Miller et al. [38]

n = 100 older patients (≥70 years; mean age 84 years) from a medical center

Stratified (community or residential care), block randomization method (blocks of 12): EG = 25, Nutritional Supplement (NS = 25), EG + NS = 24, and CG = 26

n = 7, EG = 2, NS = 2, EG + NS = 2, and CG = 1

Fall-related fracture of the lower limb and being proxy for low nutritional status (midarm circumference <25th percentile)

Rosendahl et al. [41]

n = 191 older adults (≥65 years; mean age 84.7 ± 6.5), residents of care facilities

stratified cluster-randomized, with a 2 × 2 model: CG = 50, EG = 45, EG + Protein (EG + P)  = 46, and CG + P = 50

n = 28, EG + P = 5, EG = 9, CG = 5, and CG + P = 9

being dependent in ≥1 ADL

Rydwik et al. [29]

n = 96 community-dwelling elderly (>75 years)

Consecutive randomization in batches; started with the oldest individual: Nutrition (N, n = 25), EG = 23, EG + N = 25, and CG = 23

n = 31, EG = 4, EG + N = 11, N = 7, and CG = 9

Unintentional weight loss ≥5% and/or BMI ≤ 20 and low physical activity level

Sattin et al. [25]

n = 311 (n = 20 men) transitionally frail elderly (70–97 years; mean age 80.9 years), residents of independent-living facilities

EG = 158 and CG = 153

n = 94, EG = 49, and CG = 44

see Greenspan et al. [22]

Seynnes et al. [35]

n = 27 older adults (≥70 years); nursing home residents

Strata based on 1RM values: High intensity (EG-HI, n = 8), Low intensity (EG-LI, n = 6), or CG = 8

n = 5

being a nursing home resident with a knee extension <10 kg (1RM)

Shimada et al. [30]

n = 34 older adults (67–91 years, mean age 80.8 ± 6.6 years), attending a care facility

strata based on the ability to walk outdoors without help: EG-B(balance exercise = 12), EG-G (gait training = 12), and CG = 10

n = 2, CG = 1, and EG-G = 1

Living in/or utilizing a day-care program; subjects had a certification of long-term care need by the Japanese public nursing care insurance system for frailty

Timonen et al. [33]

n = 68 older women (≥75 years; mean age 83 ± 3.9 years) discharged from an acute hospital ward

EG (n = 34) and CG (n = 34)

n = 16, EG = 10, and CG = 6

difficulties in mobility and balance and tendency to fall when walking unassisted during an acute disease

Timonen et al. [32]

see Timonen et al. [33]

see Timonen et al. [33]

see Timonen et al. [33]

see Timonen et al. [33]

Timonen et al. [31]

see Timonen et al. [33]

see Timonen et al. [33]

see Timonen et al. [33]

see Timonen et al. [33]

Wolf et al. [26]

see Sattin et al. [25]

see Sattin et al. [25]

see Sattin et al. [25]

see Greenspan et al. [22]

Wolf et al. [34]

n = 200 (19% of men) community-dwelling older adults (≥70 years)

2 EGs (tai chi, TC = 72 and balance training, BT = 64) and one CG (n = 64). Randomization in cohorts of 32 (BT = 10, CG = 10, and TC = 12) for each of the first four cohorts; the last two were randomized in cohorts of 36 (12 subjects to each group)

n = 13, TC = 6, BT = 4, and CG = 3

see Greenspan et al. [22]

Wolf et al. [36]

see Sattin et al. [25]

see Sattin et al. [25]

see Sattin et al. [25]

see Greenspan et al. [22]