Table 3.

Assessment of the certainty of evidence presented for each outcome.

Evaluation of certainty							Summary of outcomes
No. of studies	Study design	Risk of bias	Inconsistency	Indirect evidence	Imprecision	Other considerations	No. of patients		Effect		Certainty
							Therapeutic exercise	Control	Relative (95% CI)	Absolute (95% CI)
Aerobic exercise. Muscular strength – Sit-to-stand test (monitoring range 6–12 weeks; assessed with Number of repetitions)
2	RCT1,5	not seriousa	not serious	not serious	seriousb	none	26	29	–	SMD 0.15 higher (− 0.48;1.27)	⨁⨁⨁◯ Moderate
Combined exercise. Muscular strength—1 RM Leg press (monitoring range 10–12 weeks; assessed with Weight Lifted)
2	RCT8,16	Seriousc	not serious	not serious	seriousb	none	39	33	–	SMD 0.56 higher (0.08;1.04)	⨁⨁◯◯ Low
Combined exercise. Muscular strength—1 RM chest press (monitoring range 10–12 weeks; assessed with Weight lifted)
2	RCT8,16	Seriousc	not serious	not serious	seriousb	none	39	33	–	SMD 0.73 higher (0.25;1.22)	⨁⨁◯◯ Low
Balance
4	RCT1,10,29,32	Seriousa	not serious	not serious	seriousb	none	54	47	Carmeli et al10 with aerobic exercise reported significant improvements in dynamic balance between groups p < 0.05. Cai et al.29 with a combined exercise plan showed significant increase of dynamic and static balance		⨁⨁◯◯ Low
Aerobic exercise. Gait
1	RCT13	Seriousa	not serious	not serious	not serious	none	11	11	Skiba et al. found an improvement in the space–time parameters, angular changes of the limbs, and mean values of the angular deviations at the joints		⨁⨁⨁◯ Moderate
Combined exercise. Functional Fitness—6 min walking distance (6MWD) (monitoring 14 weeks)
1	RCT28	Seriousa	not serious	not serious	Not seriousa	none	6	6	Perrot et al. showed significant improvements in functional task using combined exercise via nintendo wii		⨁⨁⨁◯ Moderate

RCT Randomized clinical trial.

^aUnclear risk of bias in masking.

^bWide confidence intervals that pass the line of no effect.

^cUnclear risk of bias in 4 of 7 criteria.