Strength gains by motor imagery with different ratios of physical to mental practice

The study investigates whether high-intensity resistance training sessions aimed at maximizing neuromuscular activation can be partially replaced by motor imagery of maximal isometric contractions (IMC) without substantially reducing strength gains. Building on evidence that motor imagery engages neural systems similar to physical execution and can enhance strength, the authors compare different ratios of physical maximum voluntary contraction (MVC) training and IMC (3:1, 2:2, 1:3 IMC:MVC proportions) against an MVC-only condition and a within-subject control (untrained exercises). The primary research question is: What proportion of physical to mental practice maintains strength gains comparable to exclusive physical training in isometric tasks after a preparatory submaximal training phase? The hypothesis is that IMC-integrated protocols will yield strength gains similar to MVC-only training, reflecting neural adaptations in the early phase of strength training.

Prior work demonstrates mental training benefits beyond cognitive tasks, notably in strength contexts, implicating central neural mechanisms. Yue and Cole (1992) showed IMC can produce significant strength gains (IMC 22%, MVC 29.8%) with contralateral transfer, supporting central adaptations. Subsequent studies reported IMC-induced gains: small muscle groups (finger abduction) around 22–35% (Yue & Cole, 1992; Smith et al., 2003; Ranganathan et al., 2004) and larger muscle groups with variable effects (elbow flexion 13.5%: Ranganathan et al., 2004; plantar flexors net ~22% after controlling for control gains: Zijdewind et al., 2003; bench press ~5.7%: Reiser, 2005). Neuroimaging/EEG findings show overlapping activation during execution and imagery (PMC, SMA, M1, basal ganglia, cerebellum) and a proportional relationship between force magnitude and brain signal amplitude, supporting a central origin of IMC effects (Lotze et al., 1999; Dai et al., 2001; Ranganathan et al., 2004). Clinical evidence suggests imagery mitigates immobilization-related strength loss (Newsom et al., 2003). However, translation to applied strength training (e.g., heavy-load sessions) and optimal IMC-to-MVC ratios remains unclear. The present study addresses this gap, particularly considering a prior standardized training phase and practical exercises.

Study design: After a 4-week standardized submaximal resistance training period (twice weekly, dynamic 15RM loads adjusted over sessions), participants undertook a 4-week experimental phase (12 sessions; 3 sessions/week) of isometric training with different proportions of physical MVC and motor imagery (IMC). Four groups: M75, M50, M25 (numbers denote percentages of mental trials; 3:1, 2:2, 1:3 IMC:MVC) and M0 (MVC-only). All groups performed the same total number of trials (physically and/or mentally). Within each subject, two of four exercises were trained (bench press, leg press, triceps extension, calf raise), while the other two served as within-subject control (CO; untrained). Measurement time points: Pretest, Posttest 1 (immediately after intervention), Posttest 2 (1 week after cessation). Participants: 43 healthy sport students (20 female), mean age 22.7 ± 2.3 years, with prior strength training experience (~2 h/week) but no recent isometric high-resistance training. Imagery ability screened using the Movement Imagery Questionnaire (MIQ); inclusion required mean score >3.0 (1=very easy to image, 7=very difficult). All provided informed consent; procedures complied with the Declaration of Helsinki. Exercises and setup: Four isometric exercises differing in extremity (upper/lower) and complexity (single/multi-joint): bench press (upper, multi) via multipress with chains/carbines fixing bar at ~90° elbow flexion; leg press (lower, multi) on 45° sled fixed isometrically with knee ~100°; triceps extension (upper, single) on preacher bench with cable fixed to achieve 90° elbow angle; seated calf raise (lower, single) with thigh secured under pad. Test contractions: two 5-s MVCs per test separated by 90 s; instruction to ramp to peak over ~2 s. Measurement device and recording: Isometric linkage with steel chains and a calibrated force sensor (System DigiMax, mecha Tronic GmbH). Sampling at 1000 Hz; second-order Butterworth low-pass at 6 Hz to remove high-frequency fluctuations. Peak force from two trials used as MVC. High test–retest reliability at each time point (r > 0.93 across exercises). Training protocols: Experimental MVC sessions: four series of two maximal 5-s isometric contractions; 10 s rest between contractions, 90 s between series. IMC sessions mirrored this timing exactly. IMC instructions emphasized vivid kinesthetic imagery of maximal contraction with full muscular relaxation. Temporal cues: “position,” “start,” “OK” for onset/offset; subjects typically closed eyes. Visual monitoring ensured no overt contractions; EMG was not used, so minimal background EMG cannot be excluded, though prior work indicates near-zero activation during IMC. After each IMC session, participants rated imagery vividness (1–5; adapted from MIQ); a pretraining session familiarized them with IMC. Standardized pre-intervention training: 4 weeks, twice weekly dynamic training of two exercises at 15RM; multiple-repetition protocol to set loads; progressive increases recorded to quantify gains and familiarize participants, reducing retest/learning effects. Statistical analyses: Because individual exercises did not differ in MVC gains at Posttest 1 [F(3,84)=0.812, p=0.491] or Posttest 2 [F(3,67)=0.215, p=0.886], average percentage gain across the two trained (and two untrained) exercises per subject was analyzed. For training effects (trained vs control within-subject), 2×2 ANOVAs with between-factor Group (M0, M25, M50, M75) and repeated measures for practice condition (trained vs untrained) were conducted separately for Pre→Posttest 1 and Pre→Posttest 2. A 2×2 ANOVA (Group × Time: Posttest 1 vs Posttest 2) assessed changes between posttests. Significance α=0.05; effect sizes η² and Cohen’s d reported; data as mean ± SE. Exploratory correlations examined relationships between initial strength and gains, gender differences, and imagery vividness versus strength gains.

- Pre-intervention standardized training (4 weeks, submaximal 15RM): significant strength increases across sessions, t(85)=9.04, p<0.001, d=1.00; similar adjustments across groups (21.3%–29.6%), F(3,82)=0.49, p=0.69. - Main intervention effects (trained vs control exercises): significant within-practice effects indicating specific training gains. • Posttest 1: F(1,39)=10.46, p=0.002, η²=0.21. • Posttest 2: F(1,31)=5.83, p=0.022, η²=0.16. - Group outcomes (mean percent MVC gains): • Posttest 1: M0 (MVC-only) ≈ 4.3%; IMC groups (M25, M50, M75) ≈ 3.0–4.2%. • Posttest 2 (1 week later): M0 ≈ 8.3%; IMC groups ≈ 2.6–4.0% (gains remained stable; no delayed increase). • Earlier summary indicated CO (untrained) ~ −0.2%. - No significant group main effects or Group × Practice interactions at either posttest: • Posttest 1: Group × Practice F(3,39)=0.27, p=0.85, η²=0.14; Group F(3,39)=0.004, p=1.00, η²=0.00. • Posttest 2: Group × Practice F(3,31)=0.57, p=0.64, η²=0.05; Group F(3,31)=1.68, p=0.192, η²=0.16. • Group × Time (Post1→Post2) not significant: F(3,31)=1.63, p=0.202, η²=0.137. - Initial strength vs gains: negative correlations (weaker at baseline tended to gain more); significant only for calf raise, r=−0.61, p=0.005 (n=20). Other exercises r=−0.15 to −0.28 (ns). - Gender: no difference in IMC groups’ gains (females 3.6% vs males 3.5%), t(32)=0.013, p=0.99. - Imagery vividness: positive association with strength gains. • Posttest 1: r=0.30, p=0.041 (n=34). • Posttest 2: r=0.40, p=0.019 (n=27). • Median split (good vs excellent imagers): greater gains in excellent imagers; t(32)=1.75, p=0.045. - Overall, integrating IMC at 25–75% of trials produced short-term strength gains comparable to MVC-only training, though MVC-only showed a larger additional increase after one week without training.

Findings support the functional equivalence between motor imagery and physical execution for neural adaptations in early-phase strength training. After a preparatory submaximal phase, substituting a portion (25–75%) of high-intensity isometric training with IMC yielded specific strength gains in trained exercises comparable to MVC-only immediately post-intervention. This indicates that central neural processes engaged during IMC can drive meaningful improvements in maximal isometric force. One week after cessation, MVC-only showed a further increase (potentially reflecting delayed peripheral adaptations), whereas IMC groups’ gains remained stable, suggesting IMC primarily targets neural mechanisms without additional delayed muscular contributions. Inter-individual variability was notable; imagery vividness moderately predicted strength gains, implying that the quality of internal simulation influences training efficacy. The smaller IMC effect sizes compared to some prior studies likely reflect higher pretest performance due to the initial 4-week training and the relatively short total IMC duration, as well as the isometric, task-specific design. Collectively, results address the research question by showing that a substantial proportion of high-intensity sessions can be replaced by motor imagery without significant loss in strength gains in the short term, with potential practical benefits for managing neuromuscular load.

High-intensity isometric strength training sessions can be partially replaced by motor imagery of maximal contractions (25–75% of trials) with minimal reduction in short-term strength gains in recreationally trained individuals. Gains from IMC-integrated protocols are specific to trained tasks and remain stable over a subsequent week, whereas MVC-only may exhibit additional delayed improvements. Imagery vividness relates positively to strength outcomes, highlighting the importance of imagery quality. IMC is a viable supplementary method to optimize neural activation while potentially reducing musculoskeletal stress. Future research should: (1) examine effects in highly trained athletes; (2) test dynamic strength tasks and transfer to sport performance; (3) manipulate total IMC volume/duration to optimize dosing; (4) use physiological monitoring (e.g., EMG) during IMC; and (5) explore individual differences (imagery ability, baseline strength) to personalize IMC integration.

- No EMG monitoring during IMC; minimal background muscle activity cannot be entirely excluded despite visual checks. - Total IMC duration per session was relatively short to match MVC timing; longer IMC exposures might yield larger effects. - Sample comprised recreationally trained sport students; findings may not generalize to elite athletes or clinical populations. - Isometric-only exercises and testing limit generalizability to dynamic strength tasks. - Moderate sample size with attrition to Posttest 2 (35 completed), potentially reducing power for between-posttest comparisons. - Within-subject control used untrained exercises; no separate external control group training another modality during the intervention phase.