The tracking and optimisation of training load (TL) has recently grown in popularity in the sporting world as a strategy to decrease injury risk while improving fitness and performance. ¹ This method encompasses a wide variety of tools to measure TL, which can be organised into external and internal loads. ¹ External load (EL) is defined as “any external stimulus applied to the athlete that is measured independently of their internal characteristics”. ¹ Measures of EL include variables such as distance run, volume of weight lifted, or number of accelerations as measured by devices such as Global Positioning System (GPS) units. ^{1, 2} Internal load (IL) is any load that is “measurable by assessing internal response factors within the biological system, which may be physiological, psychological, or other”. ¹ Variables of IL include heart rate or ratings of perceived exertion (RPE). ^{1, 2} These variables can be measured using a variety of methods, such as a total value across one or multiple weeks or the change in values between weeks. ²

Rapid changes in EL and IL variables can be indicative of future injury risk or performance change. For example, Piggott ³ demonstrated that 40% of injuries in Australian Football League (AFL) players followed a change in TL compared to the previous week, while high running distances occurring over three weeks (73,721 – 86,662 meters) increased injury risk in the same population (odds ratio [OR] = 5.5). ⁴ TL has also been associated with performance changes. A study of rugby league players found that high IL, as measured by session RPE, was significantly ( p-value = 0.04) related to decreases in agility performance. ⁵ Though commonly used in sports, programs that optimise TL to mitigate injury and performance loss may be of benefit in other populations. Tactical populations are one such group that may see benefit from optimising load prescription given their high prevalence of injuries. ⁶

Tactical agencies, including law enforcement, employ periods of training within academies, which combine classroom lectures, physical training sessions, and occupational skill development. ^{7– 9} Recruits who participate in academy training are often drawn from the general population, and academy life can represent a significant increase in both physical and mental stress. ^{10, 11} This rapid increase in stress is one of the major contributing factors to musculoskeletal injuries suffered by recruits. ^{10, 11} These injuries have multiple second order effects, such as personnel and financial costs for organisations. For example, injuries to recruits can result in separations or having to leave the academy, thus a loss of qualified personnel. ¹² In addition, injured recruits elicit treatment and lost training time costs as well as costs associated with longer term care. ¹³ Given average costs to train a law enforcement recruit, with examples ranging from circa $64,000 USD ($104,000 USD when adjusted to 2023 values) ¹⁴ to $100,000 USD ($150,000 when adjusted to 2023 values), ¹⁵ injuries leading to separations can result in significant financial cost. Additionally, one of the biggest predictors of injury risk is having previously sustained an injury. ^{16, 17} Therefore, reducing injuries during academy training may ensure healthier, and potentially longer, careers for tactical personnel.

Research in military training has shown that training injuries are often overuse in nature, and may be the result of overtraining (e.g., excessive workloads). ¹⁰ Additionally, military recruits who complete high running distances (i.e., an EL variable) during an eight-week boot camp (>25 miles (40.2 km)) have been shown to be at an increased risk of injury with no further improvement in fitness. ¹⁸ While law enforcement and military encompass differing occupational tasks, law enforcement academies typically employ a similar physical training style, often noted as paramilitary training. ¹⁹ These training programs typically include body weight circuit training and long distance runs, ^{20, 21} and are often considered to be a contributing factor to injuries in law enforcement recruits. ¹⁹ Crucially, it is still vital for these professions to engage in physical activity to improve physical fitness as various components of fitness have been found to significantly relate to occupational task performance. ^{22– 24}

Due to the impacts that injuries have on law enforcement, at both the individual and organisational level, it is vital that TL-related contributors to injuries are identified. Identifying these contributors can help inform specific and effective injury mitigation strategies. Though previous research in sports has explored the potential relationship between TL and injury, ^{2, 4, 5} little research has been performed within the tactical environment. Therefore, the aim of this study was to examine the relationship between injury risk and TL in a tactical population undergoing academy training to help inform injury reduction and training programs. It was hypothesized that high amounts of distance covered, and large weekly changes would contribute to injury risk in this population.

A mixture of prospective and retrospective methods was used to collect data for this study. This previously published methodology ²⁵ was performed to validate a desktop analysis. The desktop analysis allowed for measuring variables of interest (e.g., distance covered per week, weekly change in distance), and applying to multiple classes ensuring adequate sample size and power. One class was followed prospectively to validate a desktop analysis which was then applied retrospectively to six other recruit classes. Full description of the methodology has previously been published. ²⁵ The adequate sample size allowed the use of statistically valid modelling procedure to accurately examine the impact of dependent variables on injury risk.

Data were available from 547 individuals, of which 431 (78.8%) were male and 116 (21.2%) were female, who participated in training during the research timeframe. A total of 76 injuries occurred across the seven classes, with injuries occurring most often during the beginning of the program (Week 2 to Week 4) and a second spike occurring around Week 13 ( Figure 1). Of these injuries, 23 (30.3%) occurred in female recruits, and 53 (69.7%) occurred in male recruits. This represents approximately 19% of female recruits and 12% of male recruits suffering an injury.

Figure 2 shows the average distance covered per week across the seven classes. Recruits covered approximately 15 to 23 km per week, except for four weeks (Weeks 1, 18, 21, and 22). The highest distances covered per week exceeded 30 km (Weeks 15, 17, and 20) for some classes. There was a large 10 km increase in the distance covered in Week 2 compared to Week 1.

Comparisons of the GLMM models found the best fitting model to utilise weekly distance, week of training, and sex to predict injury (χ ²= 38.3, p-value < 0.001). Supplementary Digital File 1 (Extended data) details the individual models made prior to the final, presented model. Referent values for these variables were as follows: distance – 0 km, week – Week 1, sex – male. Results of the model suggest that for every 0.08 km (80 m) covered, the odds of sustaining an injury were increased by a factor of 1.08 (95% CI 1.04, 1.12). As the academy progressed, injury risk decreased per week (OR = 0.94, 95% CI 0.91, 0.98). Lastly, biological sex was another significant factor, with males less likely to sustain a musculoskeletal injury (OR = 0.55, 95% CI 0.34, 0.91). The fixed effects estimate, z-value, odds ratio with 95% CI and statistical significance of each predictor can be seen in Table 1. The addition of further variables either did not significantly improve model fit or suffered from issues of convergence or overfitting. Diagnostic checks were carried out on the final model and no issues were detected (Extended Data Supplementary Digital File 2).

The results of this model suggest that higher distances covered per week resulted in an increased probability of injury ( Figure 3). As this graph was not linear, it suggested that injury risk was compounded by further increases in distance.

The model also suggests that as the academy progresses, the probability of injury decreases. Higher distances occurring in earlier weeks are more likely to result in recruit injuries, with distances of approximately 30 km resulting in higher probabilities of injury. This probability continues to decrease as the academy training progresses ( Figure 4).

Lastly, when analysing the impact of sex on injuries, female recruits were more likely to suffer injuries than their male counterparts, with female recruits almost twice as likely to suffer an injury as male recruits when compared across similar distances and weeks ( Figure 5).

The aim of this study was to investigate the impact of TL on injury risk in law enforcement recruits. The results of this study found distance covered per week, week of training, and biological sex were significant predictors of injury. Higher distances covered per week, and earlier weeks of training both increased injury risk, while female recruits were significantly more likely to be injured compared to male recruits. The current data has important implications for the training staff of law enforcement recruits with regards to reducing injury risk in their personnel.

The results from this study indicate that decreasing the distance covered, particularly in the beginning of training, may be a method to decrease TL and therefore reduce injury risk. This follows research in other populations, such as runners and military recruits, that show an increasing injury incidence with higher running distances. ^{18, 52– 54} Research by Trank et al. ¹⁸ found that military recruits who completed over 25 miles (40.2 km) running over an eight-week period had a significantly higher injury rate than recruits who ran fewer than 25 miles (40.2 km). Further, research in AFL players suggested three-week running distances between 74 to 87 km were associated with an increased risk of injury. ⁴ Recruits in the law enforcement population informing this research covered up to 30 km per week, a similar distance to the aforementioned AFL players when averaged over a three-week period. However, a potential difference between these TL relates to the differences in the intensity at which the distances were covered. In AFL players, distances were covered mainly by running, potentially at higher speeds, with previous research identifying players ran between 6.8 and 7.5 km/hour on average during games. ⁵⁵ However, though law enforcement recruits may be working at a lower intensity they are also likely to have lower levels of fitness. Elite AFL players have been found to average VO _2Max around 60 ml/kg/min, ⁵⁶ compared to recruits in this population who have an average VO _2Max of 40.2 ml/kg/min. ³¹ Thus, while running speeds may have been slower in the law enforcement population, their relative intensity may still be high. Furthermore, previous research has shown that lower levels of aerobic fitness are associated with higher rates of injuries in both AFL ⁵⁷ and tactical populations. ⁵⁸ It has also been proposed that lower aerobic fitness may negatively impact the ability of AFL players to tolerate changes in TL and therefore present with an increased risk of injury. ⁵⁹ Thus, the lower overall fitness of law enforcement recruits, itself a risk factor of injury, ^{60– 63} may also limit the amount of TL that can be tolerated prior to injury.

The higher distances covered may be leading to higher injury risk due to increased exposure, which has previously been shown to be a determinant of injury. ⁶⁴ While decreasing total distance covered may be a valid strategy to minimise injury risk by limiting exposure, it is vital that recruits are sufficiently exposed to physical training in order to improve their physical fitness, which is crucial to the typical occupational tasks performed in law enforcement. ^{22– 24} Alternate physical training strategies that improve physical fitness, while controlling for distance covered, warrant consideration, such as interval training and Ability Based Training whereby fitter individuals run further than less fit individuals. ^{19, 26}

As the law enforcement population informing this study often engages in long distance running as a means of training, ²⁰ the use of alternative strategies to reduce distance covered may benefit injury risk, while still being sufficient to improve physical fitness. Studies have shown that interval training may reduce injury risk in military recruits, though the true effect of interval training is unknown due to presence of other injury mitigation strategies in these studies. ^{65, 66} However, although interval training may reduce the number of loading cycles, the increase in intensity and ground reaction force may impact risk of injury. ⁶⁴ The use of cross training is another method that has been proposed to limit injuries due to its ability to reduce repetitive stress on specific body parts, ⁶⁷ though research will be needed to examine its impact on injury in this population.

Utilising strength training may be another method to reduce distance covered, potentially reducing injury risk, while still improving fitness in recruits. Although this form of training will be more likely to improve aspects of muscular strength and power, these factors are still vital to the performance of occupational tasks ^{23, 24} and have been related to injury risk in previous studies. ²⁷ Applying a varied physical training program could lead to a more well-rounded fitness profile, and more importantly for the purposes of this study, reduce total distance which was a predictor of injury risk.

Week of training was also a predictor of injury, with a higher probability of injuries occurring earlier in the academy. Higher risk of injury earlier in the program may be due to an increase in TL and stress as recruits transition from civilian life. Research on college students (mean age of 22.6 years) found that they could cover distances ranging from 1.9 to 3.2 km per day, or approximately 9.5 to 15.8 km across five days (the same length of the typical work week for recruits) as part of normal routine. ⁶⁸ Recruits participating in this academy can be expected to cover approximately 20 km a week during the early stages of the academy. ²⁵ This transition and increase in TL has previously been theorised to contribute to the higher levels of injuries experienced in recruit populations. ¹⁰ The use of strategies such as ABT and population-specific periodisation may ease this transition, resulting in a more optimal TL. Law enforcement academies tend to engage in a “one size fits all” training approach which, in reality, may fit no one. This approach imposes a standardised TL across a population with varying levels of fitness and potentially overloads unfit recruits and underloads fitter recruits. ¹⁹ The implementation of ABT, where recruits are trained at a more appropriate level given their ability, may reduce the TL for recruits at a higher risk of injury while ensuring fitter recruits receive sufficient stimulus to improve fitness. ¹⁹ Likewise, a population-specific periodised approach, targeted to the occupational demands, whereby periods of recovery are incorporated, may enable a graduated increase in TL. This concept can be seen by Ross and Allsopp, ⁶⁹ who advocated for periods of strategic rest (termed “orthopaedic holidays”) that allow for sufficient recovery while not interfering with physical training. Though few weeks had distances covered greater than 30 km, it is possible that similar fitness improvements can be made while covering distances between the 16 and 24 km ranges, thus mitigating injuries related to overtraining. Although suboptimal TLs may be a factor in the relationship between training week and injury risk, this may also be explained by the healthy worker effect. The healthy worker effect refers to the idea that injury risk may decrease as academy training progresses due to the recruits at higher injury risk experiencing injuries earlier in the academy. ⁷⁰ Future research will be necessary to study the impact of implementing cross training, strength training, and ABT on the distance covered, and more crucially, the injury risk and fitness improvements. Alternatively, physical training programs or sessions provided for recruits prior to their commencement of training may ease the transition into academy.

Lastly, sex was also a significant predictor of injury with female recruits more likely to suffer injuries than males. Previous research in tactical populations, especially during periods of training, have also found that injuries are significantly higher in female personnel. ^{60– 63, 71} However, research in this area shows that this significant difference between male and female recruits is reduced when accounting for recruit fitness. ^{60– 63} As the injury disparity between sexes is reduced when fitness is accounted for, a recruit’s fitness level may be a more important predictor in a recruit’s injury risk than sex. ^{60– 63} This supposition is supported by research showing that female recruits, on average, are less fit than male recruits, ^{31, 72} and that fitness has previously been tied to injury risk across a variety of tactical populations. ^{17, 58} Despite these findings, this study was not able to adequately assess fitness (as measured by the PT500, WSTB, and associated assessments) as an injury predictor. This was due to the test models suffering from problems such as overfitting and convergence, likely to have been exacerbated because only initial values were used, meaning that each recruit had the same weekly fitness scores. As fitness is known to change during recruit training, ³¹ future research could attempt to measure fitness level at different stages of the training program.

Limitations are present in this study. Firstly, the lack of variation in the desktop analysis limits the conclusions of this model. While academy training staff do consider academy programs by cohort models, more individualised data may have provided more information on the relationships between variables of interest and injuries. The only individualised variable in the data was sex and if other individualised variables (e.g., age or weight) were available for modelling, it is unknown if sex would still be a predictive factor. The lack of an IL measure is another limitation. The use of IL is important as different individuals may respond differently to the same training stimulus and is an adequate measure of intensity. Measuring internal TL and assessing its relationship to injury risk will be a vital future avenue of research, especially when understanding the impact of physical training strategies such as interval training or strength training that will reduce the distance covered but may result in higher intensities. Future research could compare the injury rates of utilising an interval training program versus a standard physical training program more effectively if it considered not only external loads (e.g., distances covered) but also internal loads (e.g., RPE or heart rate). Additionally, though the week of training was found to be a significant predictor of injury, this may be due, in part, to the healthy worker effect. ⁷⁰ Injury risk may decrease as the academy progresses due to recruits that are likely to be injured suffering injuries and separating from the academy in the beginning. Despite the lack of individualised data measuring EL and IL being perceived as a limitation to this study, it is unlikely that law enforcement academies will have the resources to individually monitor recruits throughout the academy. For example, the average number of recruits per class, based on 547 recruits across seven classes, was 78, with multiple classes running concurrently across various locations. Supplying each recruit with a GPS and heart rate monitor device may prove to be too costly for the average law enforcement academy especially when combined with the human resource cost of needing to collect, analyse, and implement plans based on this data in real time. Academy staff are experts in the field of law enforcement but may have little experience or knowledge in the specialised field of data and sports science. Additionally, law enforcement academies rarely employ dedicated strength and conditioning or sports science professionals due to the added financial costs. Thus, noting the limitations, this research utilised a pragmatic approach that the average law enforcement organisation could implement at a cohort level (i.e., a desktop analysis with distances as the variable). Other cost-effective options, such as step counters/pedometers, may also warrant consideration to provide more individualised information. While future research may continue to explore the concept of individualised load monitoring, realistic and pragmatic strategies for these populations will be vital until individualisation is possible within the given academy constraints.