Thirty-five RCTs and quasi-RCTs were included (n=1,345).
Adults with unimpaired quadriceps femoris muscles (17 RCTs).
The mean quality score was 43% (range: 22 to 56).
NMES versus no exercise.
The funnel plot for all studies was asymmetrical, suggesting the possibility of publication bias.
Independent NMES versus no exercises (12 studies, n=235): all studies showed point estimates favouring NMES. The meta-analysis showed significant benefit with NMES compared with no exercise (WMD 8.00 Nm, 95% CI: 2.79, 13.21). Significant heterogeneity was found (P=0.0006).
Superimposed NMES versus no exercise (2 studies): one study showed significant improvement with NMES, while the other showed no significant difference between treatments. The pooled result was statistically significant in favour of NMES (WMD 25.61 Nm, 95% CI: 9.74, 41.75).
NMES versus volitional exercises (10 studies).
The funnel plot was asymmetrical, suggesting the possibility of publication bias.
Independent NMES versus volitional exercises (8 studies, n=155): the meta-analysis showed no significant difference between treatment groups (WMD -11.60 Nm, 95% CI: -24.34, 1.13). No significant heterogeneity was found (P=0.9).
Superimposed NMES versus volitional exercises (2 studies): the meta-analysis showed no statistically significant difference between treatment groups (WMD -11.11, 95% CI: -37.06, 14.83).
Results for independent versus superimposed NMES (2 studies), different frequencies of NMES (1 study), NMES in knee flexion versus extension (1 study) and NMES in knee flexion for different intensities (1 study) in unimpaired adults were also reported.
Adults with impaired quadriceps femoris muscles (18 studies).
The mean quality score was 40% (range: 11 to 67).
NMES versus no exercise.
Intervention applied post-immobilisation period (7 studies): all 3 studies presenting adequate data for display in a forest plot showed a non-statistically significant benefit with NMES compared with no exercise. All but one of the remaining 4 studies showed results favouring NMES. All of the studies had methodological weaknesses.
Intervention applied during the immobilisation period (3 studies): both studies presenting adequate data for display in a forest plot showed significant benefit with NMES; the remaining study showed no significant difference between treatments.
NMES versus volitional exercises. Intervention applied post-immobilisation period (5 studies): only one of 3 studies presenting adequate data for display in a forest plot that showed significant benefit with NMES used superimposed NMES; the other 2 studies used independent NMES. One of the two remaining studies only showed benefit over volitional exercises with medium frequency independent NMES; the results of the other favoured electromyographic-assisted volitional exercises.
Intervention applied during the immobilisation period (3 studies): one of the 2 studies presenting adequate data for display in a forest plot showed significant benefit with NMES compared with volitional exercises; the other showed no significant difference between treatments.
Results for experimental versus traditional NMES (2 studies) were also reported.