The risk factors for injuries in parachuting and load exposure in the training of Chinese paratroopers

2.1 Subjects

The study included 7230 soldiers who filled out the questionnaire and 38 of them also participated in the experiment on muscle load assessment. All participants were male and had a body height of 173.5 (SD=4.4) cm, a body weight of 64.5 (SD = 7.2) kg, and an age of 22.5 (SD = 3.2) years on average.

2.2 Field study on injury

The paratroop headquarters was well informed about the research program and responsible to organize and mobilize the soldiers for the questionnaire survey, who were in the barracks and at the moment without other tasks. There was no selection of the subjects and meanwhile freedom to participate as a principle was guaranteed. Among the recruited subjects in the survey, a sub-group of subjects was selected for the muscle load measurements during platform jumping. Methods for the study were approved by the ethics committee of the related universities. Participants in the muscle load assessment were additionally informed about the non-invasive procedure of remote EMG measurement with the surface electrode.

2.3 Questionnaire

A modified Nordic questionnaire on the musculoskeletal injury was used as a basic part [8], and additional questions related to risk factors specific to military parachuting were included [2, 9]. After discussing with orthopedists from the military hospital and filling out the questionnaire draft as a test by a group of soldiers, it was improved and revised several times and at last named as “Questionnaire on the musculoskeletal injury of military parachutists”. Directly in the barracks, with aid of slides and loudspeakers, the questionnaire was shown to the soldiers with explanation, and at last a free participation was underlined again. They then fill it out by themselves. During that time the investigators walk around and answered questions if any. Of the 7230 questionnaires given out, 6844 were returned (94.7% of returning rate) and analyzed. Because of such investigation with questionnaires, no case of death was involved.

In addition to age, military service year, and other items, the questionnaire included specially designed questions such as:

Have you ever had an accidental injury during parachuting? If yes, then:

It also included a parachute canopy and airplane to be used, door to leave, and whether a weapon was carried with during the parachuting when an injury happened.

About fatigue and injury during land-based training, it asked questions like:

What were the training exercises: platform jumping, airplane door leaving, rings, or running?

For injuries during either parachuting or training the questions included: what was the type of injury: sprains, fractures, or both of them; which body part was injured: neck, shoulder, back, low-back, knee, ankle, etc. totally of 13 body parts; whether had to be admitted to a hospital. Only the injury with the last condition was defined as a severe one.

2.4 Experiment on platform jumping with EMG

The daily used platform, with a sandpit, was in the height of 1.5 m or 2.0 m and a platform with 1.0 m was added for research purposes. Standing at the surface in a neutral position was taken as a control group. Thirty-eight soldiers were as volunteers involved in the landing assessment. As daily training required, they jumped from the platform without load per hand, landed on the sandpit, and stood stabilized with their knees bent slightly. Taking into consideration the above exercises and postures, especially landing on the sandpit in standing posture with bent knees, the 8 muscles of tibialis anterior, gastrocnemius, semitendinosus, biceps femoris, rectus femoris, rectus abdominis, erector spinae, and trapezius were chosen for EMG measurement, according to the suggestion of our orthopedist colleagues and also to the related paper [10]. At first, the skin was prepared by thinning and removing the stratum corneum with abrasives and alcohol to ensure the skin impedance as low as possible, e.g., lower than 5 k Ohms. A Biotell-44 (Glonner, Germany) system, PCM12 (Data Translation, USA), and DASYLab 8.0 software (National Instrument, USA) were used to record and analyze the EMG signal. It functions with telemetry and surface electrodes thereby without interfering with their activities.

2.5 Data processing and filtering of EMG signal

At signal processing, the general index of myoelectric amplitude was Root Mean Square (RMS). The formula was: RMS = ( ∑ i = n n + N - 1 | EMG ( i ) 2 | / N ) 1 / 2

Where N equalled to 1024, each primary myoelectricity of 1 s was calculated 1 RMS. Actually, N is the sampling rate at A/D signal conversion, here the mostly recommended sampling rate of 1024 Hz was used. It should be at least twice the highest frequency cut-off of the bandpass filter, e.g., if a bandpass filter of 10–400 Hz was adopted, the minimal sampling rate employed to store the signal in the computer should be at least 800 Hz (400 x 2), as specified by Nyquist theorem, and preferably higher to improve accuracy and resolution [11].

In order to counteract the individual differences in EMG measurements for valid comparisons in different muscles and subjects, amplitude in MVE% was used.

The formula was: MVE % = RMS ACT RMS MVC × 100 %

Where RMS_ACT was the RMS amplitude of the EMG signals recorded during the activity of landing on the sandpit, RMS_MVC was the RMS of the EMG signals recorded during a MVC.

Before that, EMG signals and force of the related muscles were measured at the same time for every subject when making a maximal voluntary contraction with certain postures, to represent muscle activity in MVC% (percentage of the maximal voluntary contraction, a term in biomechanical area) and in MVE% (referring to the electrical activity during a maximal voluntary contraction, a term in physiology). During the measurement they were asked to hold a maximal contraction for approximately 3 sec, repeated 3 times with a 2-min break, the maximal muscle electricity value was used for EMG normalization. More detailed measurement and analytic methods can be found elsewhere [12–14]. For the experiment platform jumping with EMG and postures for MVE%, some pictures and original EMG signals can be found in the supplement. It should be noted that both MVC% and MVE% were used in the related context. At last, the platform height was taken as the independent variable and MVE% as the dependent one.

2.6 Statistic analysis

For statistic analysis, data from the epidemiological survey and EMG measurements were inputted to build up an Excel database and dealt with SPSS. Chi-square test was used for the injury rate analysis, ANOVA, and t-test for comparison of the EMG data, with a significant level set at 0.05. At last logistic regression was performed to select the variables according to their contribution. Taking both intensity and time into consideration, jump-years (J-yrs), as a product of the number of parachute descents and military service years, was used as a complex indicator of the exposure to parachuting. Either injury per 1000 jumps or injured persons per 100 soldiers were calculated as injury prevalence.

3.1 Injury rate

The overall injury rate for parachuting was 13.9 injuries per 1000 jumps and 24.5% based on persons, and the severe injury rate was 5.5/1000 jumps and 12.4% based on persons respectively. During the land-based training, the injury rate increased significantly to 60.7% for overall injury and 28.9% for severe injury (Chi-square test, between training and parachuting, P < 0.01).

Table 1 showed that the parachuting injury rate among exposure groups was different significantly (by Chi-square test, P < 0.05) and showed an increasing tendency along with the exposure levels (Cochran-Armitage trend test, P < 0.0001), in other words, a dose-response relationship between parachuting and injury.

However, the injury rate decreased with the exposure level, showing a reversed exposure-response relationship, if the rate was based on 1000 jumps (Supplementary Table 1).

In addition, the training injury rate of 62.6% in the fatigue group was significantly higher than that of the non-fatigue group (38.6%) with χ ² = 117.71 and p < 0.05. The injury rate increased steadily from 43.1% to 90.3% and significantly with χ ² = 404.38 and p < 0.05 when comparing it within the groups of fatigue in different degrees.

With an increase in body weight, the injury rate increased significantly, showing a dose relationship between them (Table 2). However, if compared with that in groups of body height, the injury rate was not significantly different as showed by a Chi-square test with χ² = 5.02 and p = 0.171 (Supplementary Table 2).

The high wind speed of the landing zone was the primary factor in the injuries, and it accounted for 24.7%. Furthermore, 21.3% of injuries occurred while the paratroopers carried weapons at landing. The third factor was improper parachute operation, which included dragging by parachute cord, colliding with the door of the plane, and two canopies being entangled with each other. The fourth factor was due to uneven terrain and that made parachutist implement landing techniques improperly, such as landing with legs apart, single-leg landing, and improper attempt landing (Table 3). The rest of the factors were ground obstacles, wearing protective equipment, and improper guide for landing, which accounted for 7.8%, 4.5%, and 2.6%, respectively. If comparing the different phases of parachuting, injuries accounted for 86.2% at landing, 7.6% at plane leaving, and 6.3% in altitude.

As analyzed with logistic regression, uneven terrain, high wind speed, carrying weapons, and taking a certain type of plane were selected among the multiple factors as more important ones, in such order to contribute to the risk of injury, as shown by their β coefficient estimate. On the other hand, ambient temperature, protective equipment, door to leave and body weight had no significant influence on the injury, and the factor of body weight even became negative (Table 4).

In both training and parachuting, most of the injuries were sprains. Comparing parachuting to training, fractures increased whereas sprains decreased significantly, and both types of them also increased (Table 5).

As compared from training to the period recruit before, the injuries took place mainly at the leg, ankle, and hip/thigh. When considering parachuting, it was only the injury at the ankle with a significant increase. In addition, the low-back injury proportion was also at a high level, although there was no significant difference between periods (Table 6).

With standing on ground surface in a neutral posture as a reference, the muscle electricity in MVE% increased gradually with platform height, showing a height-dependent relationship. Furthermore, the muscle electricity increased to a quite high level for all the 8 measured muscles, where MVE% of the muscle gastrocnemius, biceps femoris, rectus femoris, and rectus abdominis even reached 100% (Table 7).

4.1 Injury frequency

It was found that the overall injury rate among Chinese parachutists was 13.9 injuries per 1000 parachute jumps from this investigation. Studies in Norway [15], Australia [16], the United Kingdom [5], the United States [17], and Israel [18] demonstrated an injury rate of 11.1, 7.1, 4.5, 8.0, and 6.3 injuries per 1000 jumps, respectively. Certainly, it is difficult to compare these results directly, because of the differences in parachuting activities, definition of injury, and research method [6]. If a severe injury was defined as one that needed to be treated in the hospital, the injury rate of the British airborne troops was 2.2 per 1000 jumps [19]. In the Chinese paratroopers, such a severe injury rate was found to be 5.5 per 1000 jumps, and the overall injury rate based on population was 24.5%. Therefore, a basic impression was still available that the injury rate in Chinese parachutists was at a relatively higher level, and more effort was needed for the health protection of the soldiers. From an epidemiological point of view, this study showed a more causal relationship between exposure and response with the result that the injury risk increased with an increase in the product of parachute descents and military service time (parachute jump multiplies service year, Table 1). To the best of our knowledge, there is no report on the exposure-response relation of parachuting injury, especially in a large-scale epidemiological study. With the increase in the training intensity to a threshold, the injury rate increased to 200–300%, whereas the profit increased to 10% only [20]. It appeared that such a critical point existed also in the current result. When the exposure level was higher than 63 J-yrs, injury rates increased dramatically, which suggested that the soldiers exposed to that level should be protected as the main population. For risk assessment in parachutists, it has to be determined further at which level the injury frequency is “safe” or acceptable [21].

Furthermore, with the J-yrs increasing, the injury rate based on population was increasing, but the rate per 1000 jumps decreased correspondingly (see Table 1 and Supplementary Table 1), showing a different dose relationship surprisingly. It suggested that the so-called training effect might be involved in this injury indicator and it is so strong that it has changed the response direction. It was obvious that the differences between the injury parameters should be studied further.

Talking about study methods, the investigation on subjects in the field with questionnaires should be considered as a basic one where the accident frequency based either on a person or on jumps is available. And it is needed in some situations when there is no registration system or more detailed information is intended in an epidemiological study. For that purpose, a basic questionnaire like Nordic is also needed for parachuting injury [8]. Even when registered data are available, the quality of registering should be noted. In a retrospective audit in Australia, 21.5% (152 of 706) of descents had incomplete follow-up because of missing medical records [22]. An analysis on fall injuries using registered data in the Army Safety Management Information System in US between 1994–2002 found, description of injury causes was not completed for one-third of the accident reports [23].

4.2 Risk factors

The parachuting injury accounted for 86.2% at landing, which was in accordance with the findings of 87%, 83.8%, and 90% by investigations in the United States [24], Denmark [25], and Israel [18]. To reduce the landing injury, so-called low porosity round canopy and parachutist ankle brace (PAB) were used and found to be effective [26–28]. However, there were no such types of equipment for Chinese soldiers until now. The current results might serve as an example to support the use of PAB or a special canopy.

Furthermore, the findings of high wind speed, uneven terrain, weapon carrying, etc. such risk factors in our investigation were in accordance with the previous studies [29]. Even a critical wind speed was calculated to be no more than 6.56 m•s-1 and 3.47 m•s-1 for descents on the day and at night respectively [9]. However, the injuries due to improper landing techniques were significantly less than those reported by Ekeland [15]. Body weight as a risk factor was also reported by studies in Belgium [22] and Australia [30]. It was interesting to note that the body height was not associated with the injury rate. It suggested that maximal body weight should be established as a threshold for medical screening of parachutists. A restriction on the operating weight with extra equipment should be considered also [31]. The landing impact with heavier body weight on the muscles could be explained by our EMG measurements (Table 7) and discussed later in detail.

By using logistic regression, multiple factors could be taken into order according to their coefficient β. Clearly, uneven terrain was the most important risk factor shown by its contribution (Table 4) in the current results, but the carriage of equipment was found to contribute most to the injury rate in the study by Lillywhite [5]. It seems also logical, that landing, uneven terrain, high wind speed, carrying weapons, and heavier body weight are all related to a critical point i.e., large impact at landing for the parachuting soldiers. In addition, the impact should be also considered together with landing strategies, e.g., performing a sideways fall to reduce the reaction force. The body weight was significant in the single factor analysis but not anymore in multiple analysis and its β value even became negative. It is also possible when considering the nature of multiple factor comparison of logistic analysis. Both of them are important and not repellent with each other, according to the statistics principle.

4.3 Types and sites of injury

Among the US military HALO (high altitude-low opening) parachutists, the injury type of fractures was 35.5% and sprains/strains was 34.7% [32]. Of amateur parachutists in Denmark, 36.9% were soft tissue lesions, and 63.1% were fractured [25]. Only 14% of the injured soldiers suffered from severe injuries (fractures, knee ligament ruptures) in Norwegian paratroopers [15]. The injury type of Chinese paratroopers was similar to that of Norwegian soldiers, with lower fracture composition (10.3%), but higher sprains (80.9%). It seems that the injury type is related to their activities and experience [33]. The sites most commonly injured in Chinese soldiers were ankle and then low-back. The injury composition of the ankle increased significantly when compared in different periods, from recruitment before (11.3%) to training (15.1%) and parachuting (21.3%). A similar result was found by other studies [34], for example, in Israel paratroopers injured ankle and spine accounted for 35.6% and 14.5% respectively [18]. This predominant injury of the ankle in parachuting soldiers was very significant as compared to workers whose injury was mainly located in the low-back in any cases [35]. It might be related to the impact at landing, where the reaction force at the surface was found to be as high as 13.7 times of body weight [36]. In addition, it is interesting to note that landing impact is also an important risk factor for injuries insports [37].

4.4 Training fatigue and injury

These results showed that fatigue was very common and contributed to the injury, during land-based training. It was confirmed again that muscle fatigue was related to musculoskeletal injury in the area of occupational health and fatigue fracture in orthopedics. In the questionnaire, fatigue was generally defined only as a subjective feeling. In contrast, in the experiment on platform jumping the EMG measurement was objective and the result was surprising. The EMG amplitude in MVE% at muscles of the gastrocnemius, biceps femoris, rectus femoris, and rectus abdominis even reached 100%, which means the muscle load by platform jumping far beyond soldiers’ maximal muscle capacity. The muscle load on average should not exceed 10% MVC and never exceed 14% MVC; maximal load should not exceed 50% MVC and never exceed 70% MVC in dynamic activities [38], given acceptable limits in occupational health and ergonomics. In the present study MVE% at most of the muscles exceeded 70% even jumping from a platform of just one meter. Another surprising finding was that so many muscles, including rectus abdominis, were involved in such an action of landing on the sandpit surface, showing a complicated body movement in view of physiology and anatomy. Obviously, the platform jumping, which was a dangerous training procedure, should be eliminated or replaced by other exercises [39]. For example, deep water running was a suitable activity to achieve the required aerobic fitness with minimum injury risk [6]. However, platform jumping is a traditional training exercise for parachutists at least in China, not only for fitness but also for training the sensation and skills of parachuting and landing especially. It should be put at end of the basic training (when their muscles fitted more) if the program cannot be canceled.

Although this study was already on large-scale and was able to demonstrate a dose-relationship between parachuting and injury, there were still some limitations, e.g., no medical records for injury and death were available to confirm it.

4.5 Study implications and practical contributions

It is suggested that an injury registering system, reorganization of the training exercises, and a maximal bodyweight as a threshold for medical screening should be carried out gradually. In future PAB, low porosity canopy and new landing skills were studied and used by Chinese parachutists to reduce their injuries.