Background

The combination of resistance training (RT) associated with aerobic training is ideal for the best performance [1,2,3]. With the growth of such information and the encouragement for the greater practice of physical exercise, different RT methods have emerged [4,5,6]. Within these modalities, when considering health, well-being and quality of life, there was less concern only with aesthetics or performance gains within the sport [1, 2]. Therefore, studies that evaluate variables related to exercise safety are important [7]. Studies on the incidence and prevalence of injuries are important to identify risk factors within the modality and develop preventive strategies [8, 9]. The comparison between one modality and another is also important for practitioners to choose the best and safest RT method. This study wished to verify which RT is safest in terms of injury prevalence and incidence. Also, it ascertained the characteristics of the injured subjects, the level of severity of the injuries and what definitions of injuries the available studies use.

Methods

Protocol registration

This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) recommendations [10] and was registered in PROSPERO (ID CRD42021257010). The searches were performed in the PubMed, SPORTDiscuss and Web of Science, electronic databases using the following keywords (Additional file 1): "Resistance training" or "Strength training" or "CrossFit" or “Weightlifting” OR “Powerlifting” AND "Injury" or "Injuries" or "Sprain" AND “Incidence” or “Prevalence” AND “Epidemiology” or “Epidemiological”. The last update of the database search was conducted on March 2023.

Eligibility criteria

Studies were deemed eligible according to the PICOS criteria [10, 11] (Table 1). To be included in the review, the studies had to be available as full text, and be clinical trials focusing on epidemiological aspects of injuries that occurred with RT. There was no time limit for the selection of articles. Literature reviews, case reports, editorials, letters to the editor, technical notes and articles published in languages other than English were excluded.

Table 1 PICOS framework
Full size table

Selection of studies and data extraction

The studies were independently screened by two reviewers (TTS and ESO) for inclusion. Each reviewer studied the title of each article identified through the search, followed by examination of the abstracts. Subsequently, the full text of the articles which passed the previous stages was analysed. Disagreements between reviewers were resolved by a third senior reviewer experienced in systematic reviews and meta-analyses (RO).

Data extraction

The data collected by two authors (TTS and ESO) from the articles referred to the sample size, type of resistance training, incidence and prevalence of injuries, associated factors, severity, and definition of injuries. The American College of Sports Medicine (ACSM) defines resistance training for health and fitness as “a form of physical activity that is designed to improve muscle fitness by exercising a muscle or muscle group against external resistance” [12]. Resistance or strength training is widely performed in contemporary health and fitness environments through the use of equipment such as free weights, sectorized weight machines, plate loaded machines, weighted balls, resistance bands, and body weight resistance equipment [13].

Quality assessment

To assess the quality of the studies, the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) was used. The scale has a checklist with 22 items that receive scores from 0 (does not meet) to 1 (complies). Depending on the sum of items reached by the study [14, 15] when the study fulfilled more than 80% of the criteria established in the STROBE, the study is considered high quality if 50% to 80% of the STROBE criteria are met [16]. On the other hand, if the considered study met less than 50% of the STROBE criteria, low quality if detected [16].

Results

Search results

The initial literature search resulted in 4982 studies. After reading the titles, 4899 were excluded, leaving 63 for evaluating the abstracts. Twelve studies were excluded at this stage, leaving 51 for further evaluation. 21 were selected for data extraction. Seven investigations were selected searching the references by hand. Eventually, 28 articles were selected for data extraction (Fig. 1).

Fig. 1
figure 1

Flow chart of the literature search

Full size image

Patient demographic

Data from 13.127 RT practitioners were collected. The mean age was 28.7 ± 6.4 years. Their average weekly training was 2 to 6.10 workouts per week. The generalities and patient demographic of the included studies is shown in Table 2.

Table 2 Generalities and patient demographic of the included studies
Full size table

Seventeen studies evaluated the number of injuries in HIFT/CrossFit, three in powerlifting, three in strength training, three in weightlifting, and one in Strongman. In addition, one study looked at HIFT/CrossFit and weightlifting. Overall, the incidence of injuries ranged from 0.21/1000 h to 18.9/1000 h [17, 18] and the prevalence of injuries was 10% to 82% [19, 20]. Within the HIFT/CrossFit, the mean injury was 4.2/1000 h and 52.5%, respectively. In powerlifting, the mean prevalence of injuries was 56.6% and the incidence of 4/1000 h. Strength training studies did not show the incidence of injuries, with a mean prevalence of 12.6%. The only study on strongman reported an injury incidence of 5.5/1000 h and an injury prevalence of 82%. Weightlifting practitioners had 3.2/1000 h of injury incidence and 46.2% of injury prevalence, respectively. The greatest number of injuries were located in the shoulders [21,22,23,24,25,26], followed by the back [27,28,29]. Some studies analysed factors associated with injuries, as well as their severity and cause. Of the 28 studies included, 21 had explicitly defined an “injury” in their methods (Table 3).

Table 3 Injury characteristics
Full size table

Quality assessment

The studies were evaluated using the STROBE Checklist. The range of points acquired by the studies on the scale ranged from 8 [30] to 19 [31,32,33]. Five studies were classified at level A, 21 at level B and two at level C.

Discussion

This study investigates the injury rate among resistance training partitioners. Traditional strength training showed a lower injury rate, unlike Strongman, which was the RT method with the highest injury rate in the selected studies. In general, the reported injuries are of high severity, with shoulders and back being the most commonly affected anatomical areas. The injuries definitions were different between the selected studies.

Within sports, there is a particularity that makes it difficult to characterize an injury. Sport, unlike other contexts, makes the athlete or practitioner continue their training or participate in some competition even with pain or loss of function. Therefore, the simple absence from training or competition cannot always be characterized as an injury [34, 35]. With this in mind, most selected studies characterised the injury as any pain or change in performance within the training modality and exercises performed. Other studies were less stringent, and only considered injuries when the subject did not practice for some time. This agrees with the definition of a sport injury as a pathological process that interrupts training or competition and can lead the athlete to seek medical treatment [36]. There is a perceived difficulty in standardizing the definition of injuries in studies. No matter how difficult it may be, this must include within its definition the inability to perform the sport [34].

Traditional ST presented the lowest prevalence of injuries, at an average of 13%, demonstrating the safety of the practice of traditional ST. The low incidence (< 1/1000 h) indicated the safety of the practice [4]. The safety of traditional ST can also be explained by the different profiles of the training method [37, 38]. While other RT modalities put a greater focus on the task and constant challenge related to performing complex movements at higher intensities, traditional ST mostly focuses on specific muscle contraction [38]. Powerlifting had a low incidence of injury, very similar to HIFT/CrossFit and weightlifting [39, 40]. Powerlifting usually occurs from the high loads used in deadlift, squat and bench press [41]. Using high loads requires excellent technique and reduces the chances of injuries [42, 43]. Most of the studies identified on HIFT/CrossFit, with an average of 4.22 lesions per each 1000 h of exposure. Even with a low average, two studies showed a high rate of injury incidence [18, 44]. Szeles et al. evidenced an incidence of 18.9/1000 h lesions, well above the others [18]. This difference of almost 5 times the mean value can be explained by the different methods used to define an injury. The main justification is the non-standardization of the definition of injury. In this review, for example, seven studies had no definition of injury. Furthermore, many studies have different definitions, which increases the subjectivity of the interpretation [4]. Hak et al. found almost double the prevalence of injuries [45]. In one of the first epidemiological studies of HIFT/Crossfit conducted online, the online questionnaire, depending on how it is disseminated, may be biased towards the target audience of the survey [46, 47], as a study of injuries in a sport can draw more attention to subjects who have already had an injury. Studies with higher injury prevalence often define injuries as any pain or loss of function that makes the subject change training or results in a reduction in training performance. Other studies with lower rates have less stringent definitions with a lower degree of rigidity or no definition at all. This further increases the importance of standardizing the method of studies [48, 49].

The injury rate in weightlifting is similar to HIFT/CrossFit. However, a smaller number of studies were found, which makes it more difficult to consider fewer results as accurate as the HIFT/CrossFit. A systematic review showed values similar to those of this review [50]. The severity of injuries in this study varied greatly, and this may occur because some accidents take place during training [50, 51]. The highest prevalence found in the studies was that of Strongman [52], a sport in which athletes perform with high loads and varied movements. Specific training is responsible for increasing the chances of injury by 1.9 times when compared to traditional ST [52].

Most injuries occurred in the shoulders, followed by the back. These results are in line with previous studies in HIFT/CrossFit, weightlifting and powerlifting [46, 50], given the high loads and large ranges of motion [50, 53, 54]. It is necessary to have good stability of the scapulothoracic complex to allow less overload on the glenohumeral joint. Lower trapezius and serratus anterior activation are critical in overhead movements [47]. A single training method altered the pattern of shoulder and back injuries. Only one of the traditional RT studies verified this and realized that injuries in the lower limbs probably occurred through running and jumping [19]. Most injuries were classified as moderate, but few studies included this variable in their results [27, 52, 55,56,57]. Furthermore, the small number of studies that verified the severity of injuries does not allow generalization of the results. As these sports do not involve a constant change of direction and physical contact, injuries tend to be less severe [47, 58, 59]. Most studies did not find an association between the practitioner's sex and the occurrence of injuries. Previous untreated injuries seem to predispose to new injuries. Some of these RT methods are recent, which makes their practitioners come from other sports with an injury already treated [47, 60]. Individuals who start practising HIFT/CrossFit are 3.75 times more likely to get injured in practice [60]. Athletes with previous shoulder injuries are eight times more likely to injure the area compared to athletes with healthy shoulders [25]. In the practical context, all RT methods seem safe. Strongman reported the highest rate of injuries, but only one study was included in the analysis.

Conclusions

Traditional strength training is the safest RT method, and Strongman is the least safe regarding injuries. The anatomical sites with the highest rate of injuries are the shoulders and the lumbar region. Study methods need to be better standardized to prevent discrepant and heterogeneous results.