Journal Logo


Incidence of Shoulder Dislocations and the Rate of Recurrent Instability in Soldiers


Author Information
Medicine & Science in Sports & Exercise: November 2016 - Volume 48 - Issue 11 - p 2150-2156
doi: 10.1249/MSS.0000000000001011
  • Free


The incidence rate of shoulder dislocations presenting to emergency departments within the United States is reported as 0.239 per 1000 person-years (23). This may be an underreporting of the national incidence rate, as it does not capture injuries where care was not received through the emergency department, such as sports medicine, urgent care, or primary care settings. Glenohumeral instability injuries entered into the National Collegiate Athletic Association Injury Surveillance System account for 23% of shoulder injuries in collegiate athletes, with an incidence rate of 0.12 per 1000 practice or game exposures (10). The National Collegiate Athletic Association Injury Surveillance System data are limited in that it only represents approximately 15% of collegiate athletes and is entered on a volunteer basis by athletic trainers (10). In the collegiate population of the U.S. Military Academy, the incidence for shoulder instability events diagnosed by orthopedic surgeons for the 2004–2005 academic year was 4.35 per 1000 person-years (13). The study at the U.S. Military Academy has the advantage of capturing injuries from all enrolled students, as comprehensive medical care is provided on site. Studying injuries in closed health care systems may offer the benefit of a nearly complete capture of injury data.

Members of the U.S. military are representative of a relatively young, physically active cross section of the general adult population in a closed health care system. Although shoulder dislocation rates in the military may not be generalizable to the general population, injury data may generalize to adults who participate in physically active occupations or exercise with significant upper extremity demands (17). The incidence of shoulder dislocation within the U.S. military has been reported as 1.69 per 1000 person-years (12). The army and marines were found to have the highest rates of shoulder dislocations, at 2.3 per 1000 person-years, which could be attributable to activities common to these services that require more involvement of the upper extremities than those seen in the air force (1.2 dislocations per 1000 person-years) or navy (1.1 dislocations per 1000 person-years) (12). Therefore, identifying factors related to shoulder dislocation and recurrent instability in an army soldier population could provide valuable information on risk factors for adults who participate in occupations or recreational activities with significant upper extremity demands.

In adult populations, recurrent injury or instability after an incident shoulder dislocation is reported in 13%–67% of cases (1,2,4,15,20). Younger age is often associated with higher rates of recurrence (2,4,15,20). Authors of a recent systematic review (9) concluded that good quality studies also indicate that greater tuberosity fractures may also be a key risk factor for recurrence, yet weaker evidence exists regarding other concomitant bony injuries and nerve palsies as risk factors for recurrence. Examination of the incidence of shoulder dislocation and development of chronic or recurrent injury in soldiers may further inform our understanding of injury risk in relatively young, active adults and adults in higher physical demand tactical occupations such as the military and first responders (police, firefighters, and rescue personnel). Identifying risk profiles for incident and recurrent injuries in individuals with higher demand activities is needed to inform injury prevention programs or modifying treatment approach to mitigate recurrence.

The purpose of this study was to quantify the 10-yr incidence of shoulder dislocations within active duty U.S. Army soldiers, as well as the risk of developing a chronic or recurrent injury after an initial dislocation. We calculated the incidence of glenohumeral dislocations during a 10-yr period from 2002 to 2011. In addition, we examined chronic and recurrent injury rates within 2 yr of initial dislocation. Risk factors for incident and chronic/recurrent injury were identified to develop greater understanding of this injury and prognosis.


This study was approved by the Institutional Review Board of the U.S. Army Research Institute of Environmental Medicine. Glenohumeral dislocations were identified by searching for the Internal Classification of Diseases, Ninth Revision (ICD-9) codes for glenohumeral dislocation (831.00, 831.01, 831.02, and 831.03) within a soldier’s medical encounter data. The data for this study were obtained from the Total Army Injury and Health Outcomes Database, a data repository and epidemiologic research tool used by the U.S. Army Research Institute for Environmental Medicine. The incidence of glenohumeral dislocations was calculated based on the initial appearance of a glenohumeral dislocation code within a soldier’s medical encounter data from 2002 to 2011. To ensure capturing incident cases, individuals with glenohumeral dislocation codes or chronic glenohumeral instability codes (718.81, 718.31, and 718.21) in their medical encounter data any time before the beginning of the study period (January 1, 2002) were excluded from this study. A retrospective cohort study was conducted to examine the demographic variables of age, sex, education, ethnicity, race, and marital status as risk factors for incident glenohumeral dislocation.

Chronic or recurrent injury after initial glenohumeral dislocation was defined as coding of dislocation or instability >3 months and up to 2 yr after injury. In addition to the risk factors examined for initial injury, comorbid scapular and proximal humerus fractures as well as upper extremity nerve injuries were also examined as risk factors for chronic or recurrent injury. These injuries may be identified immediately upon incident care for a glenohumeral dislocation or discovered during follow-on care for the injury, so diagnosis within 30 d of the dislocation was considered as co-occurrence. Fractures of the proximal humerus were categorized according to the anatomical location as allowed by ICD-9 code (greater tuberosity [812.03], head/epiphysis [812.09], other closed fractures of the proximal humerus [812.00, 812.01, 812.02]). Fractures of the scapula were categorized according to the anatomical location indicated by ICD-9 codes (glenoid [811.03], other fractures of the scapula [811.00, 811.01, 811.02, 811.09]). In addition to fractures, upper extremity nerve injuries were also examined as risk factors for chronic or recurrent injury (brachial plexus [953.4], axillary nerve [955.0], other peripheral nerves [955.1–9]).

Data analysis

An incidence rate for shoulder dislocations was calculated as a 10-yr rate, per 1000 person-years. Soldiers contributed person-time to the study from the time they entered the study until they either sustained a shoulder dislocation or left the army or until the end of the study. The data were structured using the Andersen–Gill method, and a Cox proportional hazards model was constructed to calculate incidence rate ratios (IRR) for the risk factors of incident shoulder dislocations. The rate of recurrence was calculated among incident cases. A logistic regression model was used to calculate odds ratios (OR) for factors related to developing chronic or recurrent injury after an incident shoulder dislocation.


During the 10-yr study period, 15,426 incident shoulder dislocation were identified among 1,261,297 active duty soldiers, with a 10-yr incidence rate of 3.13 per 1000 person-years. Figure 1 displays the yearly incidence rates. Descriptive variables for soldiers who sustained shoulder dislocations or chronic or recurrent shoulder dislocations and for those selected as controls are available in Table 1. Results from the Cox regression assessing risk factors for incident shoulder dislocation are available in Table 2. Younger age and male sex were associated with increased risk of injury. The rate of shoulder dislocations generally decreased with increasing age. The referent category for age was age >40 yr. All categories 40 yr and younger showed increased risk of shoulder dislocation compared with those older than 40 yr. The greatest increased risk was for those ≤30 yr (age <20 yr, IRR = 1.67, 95% confidence interval [CI] = 1.49–1.86; age 20–25 yr, IRR = 1.90, 95% CI = 1.73–2.01; age 26–30 yr, IRR = 1.66 95% CI = 1.51–1.82). Soldiers age 36–40 yr had only slightly greater risk of incident shoulder dislocation (IRR = 1.11, 95% CI = 1.01–1.23) than those >40 yr old. Males had 64% greater risk of incident shoulder dislocation than females (IRR = 1.64, 95% CI = 1.55–1.74). White race was used as the reference for the race variable, and all other race categories showed lower odds of sustaining a shoulder dislocation. Soldiers with advanced degrees (beyond 4-yr college degrees) showed less risk of sustaining a shoulder dislocation (IRR = 0.81, 95% CI = 0.73–0.90).

Descriptive characteristics of soldiers within this study, those experiencing an incident shoulder dislocation, and those developing chronic or recurrent injury.
IRR with 95% CI from regression analysis for factors associated with shoulder dislocations.
Yearly incidence rates for shoulder dislocations among soldiers from 2002–2011.

Chronic or recurrent instability occurred in 28.7% (n = 4328) of cases that did not undergo surgery within the first 3 months of injury (n = 15,064). Similar to the risk for incident injury, younger age showed a greater risk for chronic or recurrent injury. Contrary to the risk for incident injury, men showed a decreased risk for chronic/recurrent injury (OR = 0.80, 95% CI = 0.71–0.91). There were no statistically significant differences for chronic or recurrent injury based on marital status or educational background. When examining risk associated with the fracture and nerve injury comorbidities (Table 2 and Fig. 2), individuals who were also diagnosed with an axillary nerve injury were significantly more likely to have chronic or recurrent instability (OR = 3.64, 95% CI = 1.56–8.46). Complete results for the regression analysis for risk factors associated with chronic or recurrent injury are available in Table 3.

OR with 95% CI from regression analysis for factors associated with chronic/recurrent shoulder dislocations.
OR and 95% CI for the risk of chronic or recurrent injury based on the presence of a comorbid fracture or nerve injury are represented here graphically. Only the results for Axillary nerve injury reached statistical significance (*).


Consistent with previous reports in the medical literature, we found that younger individuals (12,23) and men (10,12,23) were at greater risk for sustaining a shoulder dislocation. The risk of dislocation generally decreased as age increased. Because all racial categories showed a protective trend compared with the referent group of white, the results indicate that white people may have a greater risk for sustaining shoulder dislocations than other races. This racially based finding is also consistent with previous literature examining shoulder dislocations within the military (12). The finding that an advanced degree is protective with respect to risk of shoulder dislocation could be due to individuals with advanced degrees performing less physically demanding tasks and, thus, have a lower physical exposure to activities that could result in shoulder dislocation injuries.

One of the novel aspects of the current study was the attempt to quantify the rate of chronic or recurrent shoulder instability using medical encounter data. More than one-quarter (28.7%) of incident cases identified within our population were associated with chronic injury (duration greater than 3 months) or recurrent injury (subsequent dislocation recorded with 2 yr of incident injury). Counter to the findings of other studies (2,15,19), this study showed male soldiers to have a 20% decreased odds of chronic or recurrent injury than female soldiers. This is not necessarily an unanticipated finding as there has been a previous report of no significant difference in recurrence rates based on sex (20), and another study (21) reported that although men appeared to show greater risk, they were also younger than the women in the study. When looking at chronic or recurrent injury risk with respect to a comorbid brachial plexus or peripheral nerve injury, it was found that individuals with an axillary nerve injury had 3.6 times greater odds of having a chronic or recurrent shoulder dislocation. This finding is counter to the theory presented by other authors (9) that an axillary nerve injury may favor prognosis because of the limited movement of the limb for a significant period after dislocation. Although the point estimate for brachial plexus injuries and other peripheral nerve injuries suggests that these increased the risk for chronic or recurrent shoulder dislocation, the CI values were wide and include values that do not support that conclusion. There were a small number of comorbid nerve injuries with respect to the overall sample size, which may have affected the variability surrounding the results of the statistical analysis. These injuries are recognized to sometimes occur concurrently with a shoulder dislocation (6,16,22), and it is possible that nerve injuries could adversely affect motor control or limb proprioception and diminish an individual’s ability to properly stabilize or protect the shoulder joint during activity. The findings of this study suggest that further investigation is warranted looking at outcomes and management strategies for patients who sustain a shoulder dislocation with concurrent nerve injuries, to include the mechanism of injury and the initial management of the injury (e.g., timeliness of reduction), as these may be associated with occurrence and severity of these injuries.

Previous studies report that comorbid fractures of the greater tuberosity of the humerus (2,7,15,20,21) are associated with decreased likelihood of recurrent shoulder dislocations, although there are mixed reports on whether glenoid rim fractures (2,14,18,20,21) or Hill–Sachs lesions (2,4,5,18) have a positive or negative relationship with prognosis. A meta-analysis by Olds et al. (9) published in 2015 using data from several cohort studies showed that being female, being older in age, or sustaining a greater tuberosity fracture appear to be key factors associated with decreased risk for recurrent instability, with weaker evidence as to whether bony Bankart injuries or nerve palsy decrease the risk for recurrence. The trends noted in the current study with respect to proximal humerus fractures leaned toward a protective effect for greater tubercle fractures, a negative effect for head/epiphysis fracture (best ICD-9 approximation of Hill–Sachs), and a positive effect from other closed proximal humerus fractures. However, no definite conclusion could be made from the results because the 95% CI included values indicating the possibility of a positive or negative effect for these variables. The findings of this study showed a trend toward comorbid scapular fractures being protective against chronic or recurrent shoulder instability. However, the 95% CI here also included values that could represent a positive or negative effect, as with the humeral fractures. Because of the variability of diagnoses that may be included under several of the fracture codes (e.g., humeral head/epiphysis fracture, glenoid fracture, and designations of “other”), it is possible that positive or negative prognostic associations of some injuries had opposing effects that negate potential findings of specific injuries coded under some of the broader injury classifications. Similar to the findings surrounding nerve injuries within this study, there were a relatively small number of fractures available for statistical analysis, with respect to the overall sample size.

Our study showed a 10-yr incidence rate for shoulder dislocations in the army that was slightly higher than that previously reported by Owens et al. (12) using data from the Defense Department. The higher rate in our study is likely due to the inclusion of all the directional dislocations with associated codes (anterior, inferior, and posterior) versus only the nondirectional code in the previous study. Before performing the current study, we did a direct comparison of methods between our study and the previous study (12) and found that our methods captured more dislocations. Analyzing our data with the previously described methods and period used by Owens et al. (12) revealed a nearly identical incidence rate. Therefore, we feel that we have accurately captured a greater number of shoulder dislocations using our methods and have not overreported.

Our recurrence and chronic injury rates of 28.7% fell within the broad range of 13%–67% reported by other authors (1,2,4,15,20), albeit toward the lower end of the range. This may be due in part to the limited recurrence time frame we set, which is no greater than 2 yr after initial injury. We did so because it is unclear whether an injury sustained greater than 2 yr from initial injury is related to structural or functional impairment from the initial injury or the individual likelihood of a particular person sustaining a shoulder dislocation. Interestingly, the 2-yr follow-up study by Hovelius et al. (3) showed a recurrence rate of around 30% in their cohort of 255 patients aged 40 yr and younger, which is very similar to the rate in the current study. The 25-yr follow-up in the cohort studied by Hovelius et al. (4) showed that the overall recurrence rate reached 57%. Because studies (1,3,4,15) have shown that the majority of recurrent shoulder dislocations occur within 2 yr of the incident injury, the 2-yr time frame used in this study appears appropriate as a conservative temporal measure to assess recurrent shoulder dislocations. This is in agreement with Robinson et al. (15) who described the 2-yr follow-up time as likely adequate to detect clinically important trends with respect to recurrent shoulder dislocations.

This study was performed using data from a closed health care system. Because all soldiers are required to seek care within this system, the data provide rather comprehensive information regarding health care encounters for soldiers. Even when soldiers are seen at facilities outside the health system’s network, data are obtained for the encounter when civilian facilities bill for services provided to soldiers. Therefore, these data are well suited to quantify various injuries or illnesses within soldiers, at the population level. There is also equal accessibility to health care for all soldiers, so this mitigates the effects of socioeconomic factors that could influence accessibility to health care services in some areas.

Although soldiers participate in some tasks and situations not experienced by the public, the majority of physical training and military occupational activities are similar to the physical activities seen in various subpopulations of young active adults within the United States. In fact, sports and exercise are a leading cause of injuries within the active duty military (8). Because soldiers are recruited from the general young adult population, this study provides useful information on the risk of shoulder dislocations and chronic/recurrent dislocations as seen in young and physically active adults (17).


The first limitation that must be expressed is that this study was completed using medical encounter data. Therefore, we lack information such as the mechanism of injury and other items of the medical and injury history. However, this type of database research is used in the medical community, and it is helpful in revealing the magnitude of injury or illness and potentially related factors within large samples or populations. There are also some potential data quality issues that must be noted. It is possible that data quality could be affected by errors in coding by health care providers or medical coders. Because these errors are likely random and could occur equally across the population, it is unlikely that there was a systematic effect on the results of this study. It must also be noted that we included posterior glenohumeral dislocations in our overall description of incidence and chronic/recurrent rates. Although the clinical management of this injury may draw a different approach than anterior or inferior dislocations, we believe it is still appropriate to capture dislocations of the posterior direction within the entirety of glenohumeral dislocations. Previous work that has examined glenohumeral dislocations used the code 831.0 (12), which is a nondirectional code that likely also captured posterior dislocations. Consequently, our inclusion of the posterior direction code simply provided consistency in examining shoulder dislocations using our data. Furthermore, the vast majority of the data were represented by the nondirectional diagnosis code, so performing statistical analysis based on the directional codes was not feasible. Another limitation of our use of ICD-9 codes to capture chronic or recurrent injuries is the inability to differentiate whether injury occurred to the same or contralateral shoulder. Because dislocation of the contralateral shoulder has been reported in other studies to range from 3.2% to 9% (3,20,21) during a follow-up period of several years, this study could have a similar percentage of contralateral shoulder dislocations captured within our definition of chronic/recurrent instability. These cases still represent individuals who had repeat bouts of shoulder instability, although it may have been with respect to the contralateral shoulder. Lastly, this study also lacks information on injuries before entry into the military, so it is possible that some of the incident cases were actually recurrent injury in some individuals. This limitation is common to many epidemiologic studies, where capturing historical information before the study is often difficult. That limitation noted, the methodology of this study was such to capture incident cases while in military service, which improves upon a limitation of prior studies using medical encounter data from military populations. Although the risk factors identified within this study would all be considered nonmodifiable or not potential targets of preventative interventions, these factors should be considered as factors associated with prognosis after injury and may help guide clinicians in advising patients on periods of refraining from potentially risky activity after injury.

Future research

This study provides information that can help guide the design of future prospective studies. There is a paucity of prospective studies on risk factors for shoulder dislocation injuries. Although a study by Owens et al. (11) provided information on physical examination findings that may predict shoulder instability injuries, evidence on modifiable risk factors for shoulder instability injuries is lacking. These studies would benefit clinical decision making and patient education/expectation with respect to managing shoulder dislocations and minimizing risk for recurrence. These data would also be useful in developing primary prevention strategies for incident shoulder dislocations and prevention strategies for recurrent instability. Large prospective epidemiological studies examining recurrence of shoulder dislocations are needed to further examine potential relationships with comorbid fractures and nerve injuries. The high rates of recurrent instability noted in this study and previous studies highlight the need for injury prevention strategies, particularly in young and physically active populations such as young athletes and military personnel. A future direction for research on shoulder dislocations using large scale population-level medical encounter data is to examine health care usage associated with these injuries, as well as chronicity, recurrence, and resultant disability according to care received (e.g., rehabilitation, specialty consultation, early or later surgical intervention, pharmaceutical use, or imaging). Understanding the amount of health care usage involved with these injuries will help health care providers and patients make decisions that guide care toward expedient recovery, minimizing patients’ time within the health care system, and decreasing chronicity and recurrence. This is particularly true if relationships are found between prognosis and management strategy or intervention.


In looking at the active adult population of the U.S. Army, the risk for shoulder dislocations appears to decrease with increasing age. Although men showed greater risk for shoulder dislocation injuries, they were actually less likely than women to have chronic or recurrent injuries. More than one-quarter of incident cases in this population became either chronic or recurrent. Upon examining factors associated with developing a chronic or recurrent injury, axillary nerve injuries and younger age were associated with increased risk.

This research was conducted at the U.S. Army Research Institute of Environmental Medicine. No external funding was used for this research. The authors have no conflicts of interest to declare.

This study was approved by the Institutional Review Board of the U.S. Army Research Institute of Environmental Medicine. The authors certify that they have no affiliations with or financial involvement in any organization or entity with a direct financial interest in this study.

The opinions or assertions contained herein are the private views of the author(s) and are not to be construed as official or reflecting the views of the Department of the Army or the Department of Defense. Any citations of commercial organizations and trade names in this report do not constitute an official Department of the Army endorsement or approval of the products or services of these organizations.

The investigators have adhered to the policies for protection of human subjects as prescribed in DOD Instruction 3216.02, and the research was conducted in adherence with the provisions of 32 CFR Part 219.

The results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Cameron KL, Mountcastle SB, Nelson BJ, et al. History of shoulder instability and subsequent injury during four years of follow-up: a survival analysis. J Bone Joint Surg Am. 2013;95(5):439–45.
2. Hoelen MA, Burgers AM, Rozing PM. Prognosis of primary anterior shoulder dislocation in young adults. Arch Orthop Trauma Surg. 1990;110(1):51–4.
3. Hovelius L, Lind B, Thorling J. Primary dislocation of the shoulder. Factors affecting the two-year prognosis. Clin Orthop Relat Res. 1983;176:181–5.
4. Hovelius L, Olofsson A, Sandstrom B, et al. Nonoperative treatment of primary anterior shoulder dislocation in patients forty years of age and younger. a prospective twenty-five-year follow-up. J Bone Joint Surg Am. 2008;90(5):945–52.
5. Kim DS, Yoon YS, Yi CH. Prevalence comparison of accompanying lesions between primary and recurrent anterior dislocation in the shoulder. Am J Sports Med. 2010;38(10):2071–6.
6. Kosiyatrakul A, Jitprapaikulsarn S, Durand S, Oberlin C. Recovery of brachial plexus injury after shoulder dislocation. Injury. 2009;40(12):1327–9.
7. Kralinger FS, Golser K, Wischatta R, Wambacher M, Sperner G. Predicting recurrence after primary anterior shoulder dislocation. Am J Sports Med. 2002;30(1):116–20.
8. Lauder TD, Baker SP, Smith GS, Lincoln AE. Sports and physical training injury hospitalizations in the army. Am J Prev Med. 2000;18(3 Suppl):118–28.
9. Olds M, Ellis R, Donaldson K, Parmar P, Kersten P. Risk factors which predispose first-time traumatic anterior shoulder dislocations to recurrent instability in adults: a systematic review and meta-analysis. Br J Sports Med. 2015;49(14):913–22.
10. Owens BD, Agel J, Mountcastle SB, Cameron KL, Nelson BJ. Incidence of glenohumeral instability in collegiate athletics. Am J Sports Med. 2009;37(9):1750–4.
11. Owens BD, Campbell SE, Cameron KL. Risk factors for anterior glenohumeral instability. Am J Sports Med. 2014;42(11):2591–6.
12. Owens BD, Dawson L, Burks R, Cameron KL. Incidence of shoulder dislocation in the United States military: demographic considerations from a high-risk population. J Bone Joint Surg Am. 2009;91(4):791–6.
13. Owens BD, Duffey ML, Nelson BJ, DeBerardino TM, Taylor DC, Mountcastle SB. The incidence and characteristics of shoulder instability at the United States Military Academy. Am J Sports Med. 2007;35(7):1168–73.
14. Rhee YG, Cho NS, Cho SH. Traumatic anterior dislocation of the shoulder: factors affecting the progress of the traumatic anterior dislocation. Clin Orthop Surg. 2009;1(4):188–93.
15. Robinson CM, Howes J, Murdoch H, Will E, Graham C. Functional outcome and risk of recurrent instability after primary traumatic anterior shoulder dislocation in young patients. J Bone Joint Surg Am. 2006;88(11):2326–36.
16. Robinson CM, Shur N, Sharpe T, Ray A, Murray IR. Injuries associated with traumatic anterior glenohumeral dislocations. J Bone Joint Surg Am. 2012;94(1):18–26.
17. Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 3rd ed. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins; 2008.
18. Salomonsson B, von Heine A, Dahlborn M, et al. Bony Bankart is a positive predictive factor after primary shoulder dislocation. Knee Surg Sports Traumatol Arthrosc. 2010;18(10):1425–31.
19. Simonet WT, Cofield RH. Prognosis in anterior shoulder dislocation. Am J Sports Med. 1984;12(1):19–24.
20. te Slaa RL, Wijffels MP, Brand R, Marti RK. The prognosis following acute primary glenohumeral dislocation. J Bone Joint Surg Br. 2004;86(1):58–64.
21. Vermeiren J, Handelberg F, Casteleyn PP, Opdecam P. The rate of recurrence of traumatic anterior dislocation of the shoulder. A study of 154 cases and a review of the literature. Int Orthop. 1993;17(6):337–41.
22. Visser CP, Coene LN, Brand R, Tavy DL. The incidence of nerve injury in anterior dislocation of the shoulder and its influence on functional recovery. A prospective clinical and EMG study. J Bone Joint Surg Br. 1999;81(4):679–85.
23. Zacchilli MA, Owens BD. Epidemiology of shoulder dislocations presenting to emergency departments in the United States. J Bone Joint Surg Am. 2010;92(3):542–9.


© 2016 American College of Sports Medicine