Abstract
igh-intensity/sprint interval training (HIIT/SIT) improves aerobic and anaerobic perfor- mance, but it is unknown if HIIT/SIT increases strength, muscle mass/size, and muscle endurance (ME). We aimed to determine if HIIT/SIT increases strength, muscle mass/size, and ME. Databases (Ovid Medline, Sport Discus, EMBASE, and CINAHL) and the gray literature (Google Scholar) were searched for original research articles investigating the impact of HIIT/SIT on strength, muscle mass/size, and ME (23 March 2025). The risk of bias (ROB) was assessed via the Cochrane ROB 2 Tool. Meta-analyses were performed when three or more randomized controlled trials compared HIIT/SIT to a common com- parator. Fifty-four studies were included (N= 1136). Twenty-five studies had a high ROB, while twenty-nine had some concerns. Standardized mean differences (SMD)
on strength, muscle mass/size, and ME (23 March 2025). The risk of bias (ROB) was assessed via the Cochrane ROB 2 Tool. Meta-analyses were performed when three or more randomized controlled trials compared HIIT/SIT to a common com- parator. Fifty-four studies were included (N= 1136). Twenty-five studies had a high ROB, while twenty-nine had some concerns. Standardized mean differences (SMD) (95% CI) of 0.16; (−0.09, 0.40), 0.33; (−0.21, 0.87) were observed for meta-analyses comparing the effect of HIIT/SIT to moderate intensity continuous training (MICT) and non-exercise controls (CON) on FFM, respectively. A meta-analysis comparing the effect of HIIT/SIT to resistance training (RT) on leg press strength yielded a SMD of−0.82; 95% CI: (−1.97, 0.33). HIIT/SIT may induce slightly greater gains than MICT and CON for FFM, while RT is likely superior to HIIT/SIT for improving leg press strength. However, the certainty of evidence is low, and 95% CIs intersect zero for all analyses. Keywords:high-intensity interval training; sprint interval training; muscle strength; muscle hypertrophy; muscle endurance 1. Introduction The American College of Sports Medicine recommends 150 min of moderate-intensity aerobic activity and two sessions of muscle strengthening exercises per week [1]. ~50.5% of individuals meet aerobic exercise guidelines; however, only ~30% of individuals meet muscle strengthening guidelines, and just ~20% meet both aerobic and muscle strength- ening guidelines [2]. The most cited barrier to exercise is “lack of time” [3], but if benefits Sports2025,13, 293 https://doi.org/10.3390/sports13090293
Sports2025,13, 293 2 of 31 associated with both resistance training (RT) and aerobic training could be achieved more efficiently, this barrier could be lifted. A time efficient exercise method is high-intensity interval training (HIIT). HIIT inter- vals last 1–4 min and typically range from 80 to 95% of maximal effort, with constant power output maintained [4]. Sprint interval training (SIT) is performed at maximal effort with intervals lasting ~10–30 s, using fixed loads, with performance usually declining during subsequent efforts [4]. HIIT/SIT induces similar or greater aerobic fitness improvements relative to MICT in less time [4], improves anaerobic performance [5,6], and results in comparable levels of fatigue and metabolic stress as RT [7,8]. Therefore, HIIT/SIT may be time efficient methods to improve both muscular and aerobic fitness. It is currently unknown if HIIT/SIT promotes increases in strength, muscle mass/size, or ME. Aerobic training is not typically considered to be an anabolic stimulus; however, some reports of hypertrophy in response to aerobic training exist given sufficient volume and intensity [9]. The two most important training variables for inducing muscle hypertrophy are training volume (reps×sets×load) and training in proximity to momentary failure (high levels of perceived effort) [10–12]. HIIT/SIT induce high levels of perceived effort [13], while SIT results in performance declines, indicating fatigue (task failure) [14]. Typical HIIT/SIT also demands a greater training volume relative to most RT interventions [7,8]. Therefore, if high enough intensities are achieved during HIIT/SIT, hypertrophy may result. Despite the high-intensity nature of HIIT/SIT, minimal work has examined the impact of HIIT/SIT interventions on muscle hypertrophy, with most studies assessing whole-body changes in fat-free mass (FFM). A narrative review on this topic [15] detailing the acute and chronic impact of HIIT/SIT on muscle hypertrophy at the molecular, cellular, muscular, and whole-body level demonstrated that HIIT/SIT may activate similar hypertrophic signaling pathways as RT [16,17] and elevate both myofibrillar and sarcoplasmic protein synthesis following acute exercise bouts [18]. The authors concluded that HIIT/SIT elicits changes at the transcriptional and translational level associated with muscle hypertrophy but did not systematically review HIIT/SIT-induced changes in muscle size. The only meta-analysis
cellular, muscular, and whole-body level demonstrated that HIIT/SIT may activate similar hypertrophic signaling pathways as RT [16,17] and elevate both myofibrillar and sarcoplasmic protein synthesis following acute exercise bouts [18]. The authors concluded that HIIT/SIT elicits changes at the transcriptional and translational level associated with muscle hypertrophy but did not systematically review HIIT/SIT-induced changes in muscle size. The only meta-analysis examining this topic [19] compared 4 weeks or more of low-volume HIIT/SIT to MICT and found no difference in FFM, but no comparisons were made to control groups, and assessments of localized muscle hypertrophy were not included. Improvements in muscle strength are regulated by neural adaptions and increased muscle physiological cross-sectional area, induced via mechanical tension, typically achieved via RT [20–22]. In practice, lifting heavier loads (a higher percentage of an individual’s one-repetition maximum (1-RM)) yields greater gains in muscle strength com- pared to lifting lighter loads for more repetitions [23]. However, strength gains are still observed following lower load RT [24–26]. Peak pedal forces during HIIT/SIT cycling are estimated to be ~200–500 N/leg [27], while peak isometric leg press force has been docu- mented to be ~2000–2500 N/leg in similar populations [28]. Therefore, peak forces during HIIT/SIT likely range from ~10–25% of MVC/1-RM. Previous work observed strength gains from RT at ~20% 1-RM [10], suggesting that loads encountered during HIIT/SIT cycling may be sufficient to induce strength gains. Muscle endurance (ME) is the ability of a given muscle or muscle group to resist fatigue when performing resistance exercise at submaximal loads and is vital for many activities of daily life and physical performance [29]. It can be assessed at any load but is typically tested at loads corresponding to 30–60% 1-RM and can be measured in absolute or in relative terms [29]. Absolute ME reflects training-induced changes in work capacity, while relative ME is a measure of performance fatigability [29]. Low-load RT improves both absolute and relative ME when tested at low relative loads [24]. Muscle strength, mito- chondrial function and content, muscle capillarization, and habituation to exercise-induced
relative terms [29]. Absolute ME reflects training-induced changes in work capacity, while relative ME is a measure of performance fatigability [29]. Low-load RT improves both absolute and relative ME when tested at low relative loads [24]. Muscle strength, mito- chondrial function and content, muscle capillarization, and habituation to exercise-induced
Sports2025,13, 293 3 of 31 discomfort are theorized regulatory factors of ME performance [29]. HIIT/SIT increases mitochondrial function, mitochondrial content, muscle capillarization and induces high levels of perceived exertion [4]. Therefore, HIIT/SIT may improve ME; however, minimal work has investigated this. Due to the barrier of time, a majority of individuals do not meet exercise guidelines, especially muscle strengthening guidelines [3,30,31]. The benefits of HIIT/SIT on aerobic fitness are unequivocal and are more time efficient than MICT [4]. However, it is unknown if HIIT/SIT can increase muscle mass/size, strength, or ME. The aim of this systematic review and meta-analysis was to determine if HIIT/SIT improves muscle strength, muscle mass/size, and muscle endurance in healthy adults. 2. Methods 2.1. Protocol and Registration This study is a systematic review and meta-analysis. The study protocol was prospec- tively registered on PROSPERO—(Can High-Intensity Interval Training Induce Gains in Mus- cle Strength, Muscle Hypertrophy, and Local Muscle Endurance)—CRD42023400067—15 February 2023, and is reported following the PRISMA guidelines [32,33] (Figure S1). 2.2. Eligibility Criteria Studies were included if they measured the effect of a HIIT/SIT intervention(s) com- pared to either a non-exercise or other exercise control group(s) on either muscle strength, muscle size/mass, or ME in healthy individuals (including obese individuals), aged 18–80, and of any sex or training status. HIIT/SIT interventions were required to span a minimum duration of 6 weeks, with no fewer than 12 training sessions completed. Only aerobic training interventions (i.e., cycling, running, rowing) with intervals lasting 10 s to 4 min at an intensity≥80% maximum heart rate (HR)/VO2peak/work-peak were included in this review. Only randomized control trials (RCTs) were included in the meta-analysis. No restrictions were be placed on the types of studies included in the systematic review. Only peer-reviewed studies of English language were included. Studies were excluded if HIIT/SIT was combined with resistance training, supplementation, or ergogenic aids. All studies assessed muscle strength and/or muscle mass/size and/or ME, reported as pre/post intervention values or absolute/percentage change (means with standard devia- tions or standard error). 2.3. Information Sources and Search Strategy Four online databases (Ovid Medline, Sport Discus, EMBASE, and
of English language were included. Studies were excluded if HIIT/SIT was combined with resistance training, supplementation, or ergogenic aids. All studies assessed muscle strength and/or muscle mass/size and/or ME, reported as pre/post intervention values or absolute/percentage change (means with standard devia- tions or standard error). 2.3. Information Sources and Search Strategy Four online databases (Ovid Medline, Sport Discus, EMBASE, and CINAHL) were systematically searched for this review/meta-analysis (14 March 2023, 23 March 2025). The search strategy was based on including (“High-intensity interval training” OR “Sprint inter- val training” OR “High-intensity intermittent exercise” OR “Aerobic interval training” OR “SIT” OR “HIIT”) AND ((“Muscle strength” OR “Strength”) OR (“Muscle hypertrophy” OR “Hypertrophy” OR “Muscle growth”) OR (“Muscle endurance” OR “Strength endurance” OR “Muscle fatiguability” OR “Fatigue resistance”)). The full search strategy can be found in Table S1. A gray literature search and citation searching was performed using Google Scholar (14 March 2023, 24 March 2025). 2.4. Selection Process Seven reviewers (L.W, J.ML, D. D, R. B, M.DF, M. A, J. L) applied the eligibility crite- ria and selected all studies for inclusion. All reviewers screened studies independently, and studies were screened in duplicate. One reviewer (L.W) screened all studies, and the other six reviewers (J.ML, D.D, R.B, M.DF, M.A, J. L) each screened approximately
Sports2025,13, 293 4 of 31 one sixth of the studies (randomly assigned). Researchers were blinded to each other’s decisions. Disagreements between individuals were resolved via a group discussion once all studies had passed through the given phase of screening (title and abstract or full text). Cohen’s kappa and percentage agreement were calculated via Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available atwww.covidence.org, Version 2) for both the abstract/title screening and full-text screen- ing phases. For the abstract/title screening the average Cohen’s kappa and percentage agreement were 0.42 and 96.4%, respectively. A percentage agreement of 68.5% and an average Cohen’s kappa of 0.43 was observed for the full-text screening phase. Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at, Version 2) was used for this review. 2.5. Data Collection Process Data Items Study details, including author(s), participant characteristics (age, training status, number of participants), training prescription (duration of intervals, intensity of intervals, rest between intervals, and number of sets), mode of training (e.g., cycling, running, rowing, etc.), control group (yes/no), measure used to quantify strength, hypertrophy or ME, and main findings of the study, were extracted in duplicate and put into a custom data file (Microsoft Excel for Mac (Version 16.87)) and stratified by outcome (i.e., FFM, local muscle hypertrophy, strength, and ME). Disagreements between individuals were resolved via group discussion. Study investigators were contacted for unreported data/additional details when unreported data were deemed pertinent to the review/meta-analysis. When data were only provided in figures and investigators were unresponsive, WebPlotDigitizer (V4.6) was used to extract relevant data. Acceptable measurement methods for strength outcomes included: 1-RM, isometric maximal volitional contraction (MVC), and isokinetic peak torque. For muscle mass/size, limb/segment/whole-body fat-free mass (assessed using dual-energy X-ray absorptiom- etry (DXA)), air displacement plethysmography (ADP), bioelectrical impedance (BIA)), muscle cross-sectional area (CSA), muscle thickness and muscle volume (assessed via ultra- sonography or magnetic resonance imaging (MRI)), or muscle fiber CSA were included. Measures were accepted for ME at any load in absolute or relative terms or as muscle fatiguability in the form of maximum repetitions, work, or volume completed.
etry (DXA)), air displacement plethysmography (ADP), bioelectrical impedance (BIA)), muscle cross-sectional area (CSA), muscle thickness and muscle volume (assessed via ultra- sonography or magnetic resonance imaging (MRI)), or muscle fiber CSA were included. Measures were accepted for ME at any load in absolute or relative terms or as muscle fatiguability in the form of maximum repetitions, work, or volume completed. 2.6. Risk of Bias Assessment The Cochrane risk-of-bias 2 tool [34] was used to assess the risk of bias for the random- ization process, deviation from intended interventions, missing outcome data, measure- ment of the outcome, selection of the reported result, and overall analysis. The Cochrane risk-of-bias 2 tool was chosen due to its rigor, widespread use within the field, and authors previous experience with this tool [35]. Four of the authors applied the risk-of-bias assess- ment (L.W, M.DF. J.ML, M.A). Each study was assessed by two independent reviewers (blinded to each other’s decisions). Disagreement was settled by a third reviewer. 2.7. Certainty of Evidence The certainty of evidence was assessed via the GRADE (grading, recommendations, assessment, development, and evaluation) quality analysis framework for studies included in the meta-analysis (Tables S2–S4) [36]. As all studies included in the meta-analysis were randomized control trials, the evidence certainty was initially set as high (study design). Certainty was downgraded if >25% of studies were deemed as having a high risk of bias (risk of bias); if there was minimal overlap in confidence intervals or considerable het- erogeneity (I 2 > 50%) (inconsistency); if major discrepancies existed between participant demographics, training interventions, or measured outcomes or if indirect comparisons
Sports2025,13, 293 5 of 31 were made (indirectness); if confidence intervals exceeded 0.5 on either side of the standard- ized mean difference (SMD) (imprecision); and if publication bias was detected (if Egger’s test reached significance). Evidence was upgraded if there was a large effect(SMD > 0.8), plausible residual opposing confounding, and the presence of a dose response. 2.8. Data Synthesis and Analysis A meta-analysis was performed when there were three or more randomized controlled trials comparing HIIT/SIT and a common comparator condition. Based on available data from included studies, three meta-analyses were performed. Post-intervention outcome means (SD) pooled between group comparisons (control as reference) were made with a Hedges G meta-analysis, using a random effects model (95% CI), with inverse variance weighting (restricted maximum likelihood estimation). A random effects model was chosen given expected heterogeneity across studies and that the target of inference extends beyond the samples contained within each study [37]. A Hartung–Knapp adjustment was made for small samples. Heterogeneity was assessed as the between-study variance (τ 2) and proportion of variance attributable to between-study inconsistency (I 2 ). A prediction interval was calculated to provide insight into the range of predicted treatment effect values on an individual level in a new study setting (after accounting for heterogeneity and within and between study variability) [38]. All post-intervention means were pooled together regardless of intervention length. Due to too few studies (n < 10), publication and small study bias were not formally explored. A ‘leave-one-out’ sensitivity analysis was performed to identify influential studies (see Figures S2, S4 and S6). All analyses were performed in R (version 4.3.2, R Core Team, Vienna, Austria) using the ‘meta’ package. After meta-analyses, the overall certainty of evidence was rated using the GRADE approach as described above (see Tables S1–S3) [36]. 2.9. Post Hoc Protocol Deviations Due to a lack of common comparator groups, for strength and muscle size/mass out- comes, additional post hoc analyses were performed. Weighted effect sizes and percentage change were calculated in Microsoft Excel (Microsoft Excel for Mac (Version 16.87)) via the formulas below [39,40]. (ES = effect size, n =
described above (see Tables S1–S3) [36]. 2.9. Post Hoc Protocol Deviations Due to a lack of common comparator groups, for strength and muscle size/mass out- comes, additional post hoc analyses were performed. Weighted effect sizes and percentage change were calculated in Microsoft Excel (Microsoft Excel for Mac (Version 16.87)) via the formulas below [39,40]. (ES = effect size, n = study sample size, N = pooled sample size, Studypre = study pre-intervention mean, Studypost = study post-intervention mean, PooledSD = pre/post pooled standard deviation.) Confidence intervals were calculated using an alpha of 0.05. WeightedES=∑ ≍ n× Studypre−Studypost PooledSD ≡ N Weighted%∆=∑ ≍ n× Studypre−Studypost Studypre ≡ N ×100% 3. Results 3.1. Study Selection 14,874 studies were retrieved from the initial database search along with 34 studies from citation searching and 4 studies from the gray literature. A total of 184 full-text articles were assessed for inclusion, with 54 studies ultimately included in this review (Figure).
Sports2025,13, 293 6 of 31 Figure 1.PRISMA flow diagram for study inclusion. 3.2. Study Characteristics Full study characteristics are included in Tables–4. Of the 54 studies included, 32 studies (41 interventions) assessed FFM/skeletal muscle mass (Table), 19 studies (24 interventions) assessed local muscle mass/size or muscle fiber size (Table), 27 studies (37 interventions) measured muscle strength (Table), and 5 studies (8 interventions) assessed ME/fatiguability (Table). Of the 54 studies included in the review, 19 studies (35.2%) included male participants only, 10 studies (18.5%) included female participants only, and 25 studies (46.3%) included both males and females. The age of participants ranged from 18–80 years old. The training status of participants varied across studies, 16 recreationally active (29.6%), 18 sedentary (33.3%), 8 untrained (14.9%), 3 trained (5.6%), and 4 of unknown training status (7.4%). Some studies of note that initially appeared to meet the inclusion criteria were excluded due to RT being incorporated into the intervention groups’ warm up [41], the use of functional HIIT exercise interventions [42,43], and the duration of the intervention/intervals [44].
Description
This systematic review and meta-analysis investigates the effects of HIIT/SIT on muscle strength, mass, and endurance.