Increased prevalence of lower extremity soft tissue injuries and surgeries in patients with anorexia nervosa and bulimia nervosa

ABSTRACT Purpose An abundance of literature exists linking eating disorders and fracture risk. However, no studies, to our knowledge, have investigated the impact of eating disorders on lower extremity soft tissue injury or surgery risk. The purpose of this study was to determine if anorexia nervosa and bulimia nervosa are associated with prevalence of lower extremity soft tissue injuries and surgeries. Methods Patients with anorexia nervosa or bulimia nervosa over 2010–2020 were identified through the International Classification of Diseases (ICD) codes in the PearlDiver Claims Database. Patients were matched by age, gender, comorbidities, record dates, and region to control groups without anorexia or bulimia. Soft tissue injuries were identified through ICD codes, and surgeries were identified through Current Procedural Terminology codes. Differences in relative prevalence were analyzed using chi-square analysis. Results Patients with anorexia had a significantly increased prevalence of meniscus tears (RR = 1.57, CI 1.22–2.03, p = 0.001) or deltoid ligament sprains (RR = 1.83, CI 1.10–3.03, p = 0.025), and patients with bulimia had a significantly increased prevalence of meniscus tears (RR = 1.98, CI 1.56–2.51, p < 0.001), medial collateral ligament sprains (RR = 3.07, CI 1.72–5.48, p < 0.001), any cruciate ligament tears (RR = 2.14, CI 1.29–3.53, p = 0.004), unspecified ankle sprains (RR = 1.56, CI 1.22–1.99, p < 0.001), or any ankle ligament sprains (RR = 1.27, CI 1.07–1.52, p = 0.008). Patients with anorexia had a significantly increased prevalence of anterior cruciate ligament reconstructions (RR = 2.83, CI 1.12–7.17, p = 0.037) or any meniscus surgeries (RR = 1.54, CI 1.03–2.29, p = 0.042), and patients with bulimia had a significantly increased prevalence of partial meniscectomies (RR = 1.80, CI 1.26–2.58, p = 0.002) or any meniscus surgeries (RR = 1.83, CI 1.29–2.60, p < 0.001). Conclusions Anorexia and bulimia are associated with increased prevalence of soft tissue injuries and surgeries. Orthopedic surgeons should be aware of this risk, and patients presenting to clinics should be informed of the risks associated with these diagnoses and provided with resources promoting recovery to help prevent further injury or surgery.


Introduction
Eating disorders are a group of syndromes that present with both psychiatric and physical symptoms, with anorexia nervosa affecting an estimated 0.6% of the United States population and bulimia nervosa affecting an estimated 2.0% [1,2].The DSM-5 diagnostic criteria for anorexia nervosa are restriction of energy intake leading to persistently low body weight, intense fear of weight gain, and disturbance in body image; for bulimia nervosa, recurrent episodes of binge eating followed by inappropriate compensatory behaviors at least once a week for 3 months, and self-assessment being unduly influenced by body shape and weight [3].Recent studies have shown that the incidence of eating disorders among adolescents has been increasing [4][5][6].Younger females and athletes have been shown to be at greater risk, with studies estimating the prevalence of eating disorders among female athletes at anywhere from 13% to 62% [7][8][9].Although less frequently diagnosed, the sequelae of eating disorders in men and male athletes have additionally been an area of investigation in recent years [10,11].Given the high prevalence of eating disorders in athletes, it is important to investigate the impact of eating disorder diagnoses on the risk of sports-related injuries.
There is an abundance of literature demonstrating that anorexia nervosa is associated with an increased risk of stress fractures [12][13][14][15][16][17][18][19].This association is typically attributed to the amenorrhea and resultant decreased bone mineral density and bone strength present in these patients [13,14,20].This mechanism is similarly noted in patients presenting with the Female Athlete Triad, or Relative Energy Deficiency in Sport, who may not be clinically diagnosed with anorexia nervosa, but experience similar energy deficiencies and pattern of bone mineral density loss and heightened fracture risk [12,17,21].
Bulimia nervosa is not typically associated with a notable reduction in bone mineral density, and studies have shown that the risk of fracture is not significantly increased in patients with bulimia [22][23][24].Beyond the increased risk of fractures in patients with eating disorders, a limited number of studies have demonstrated an increased risk of injury broadly in female athletes with eating disorders [8,[25][26][27][28][29].However, these studies typically included a small number of participants, often homogeneous in terms of demographics, geography, or sports played.They also included fractures among the injuries examined, making it difficult to isolate the impact of eating disorders on non-fracture injuries.As anorexia nervosa is associated with malnutrition and loss of muscle mass [30,31], and bulimia nervosa is associated with electrolyte imbalance and compensatory overexercising behavior [32-34], soft tissue weakness and injury susceptibility may be elevated in these patients.It is thus important to elucidate if anorexia nervosa or bulimia nervosa are associated with non-fracture soft tissue injuries or surgeries.
To our knowledge, no study has investigated the risk of soft tissue injuries or subsequent surgeries in patients with anorexia nervosa or bulimia nervosa.Furthermore, the majority of orthopedic research has focused solely on patients with anorexia nervosa, with limited analyses of bulimia nervosa.Therefore, the aim of this study was to determine if anorexia nervosa or bulimia nervosa diagnoses are associated with an increased relative prevalence of lower extremity soft tissue injuries and surgeries.Lower extremity was selected as this typically comprises the sites of weight bearing, and thus may be subject to more stress and at elevated risk for injury in patients with anorexia or bulimia for whom physiologic or behavioral predisposing factors may exist.Orthopedic surgeons may often be the first point of intervention for young female athletes or other patients struggling with eating disorders.It is important to elucidate the extent of increased risk so that surgeons can discuss these risks with patients and take appropriate steps to encourage eating disorder recovery.

Study design and setting
A retrospective, cross-sectional matched cohort analysis was conducted using the PearlDiver (PearlDiver, Inc., Colorado Springs, CO) Mariner database, a large national insurance claims database.The population queried included all patients within the dataset with records from 2010 to 2020.

Participants/Patient population
Patients with a diagnosis of anorexia nervosa (ICD-9-D-3071, ICD-10-D-F5000, ICD-10-D-F5001, ICD-10-D-F5002) or bulimia nervosa (ICD-9-D-30751, ICD-10-D-F502) were identified using International Classification of Diseases (ICD)-9 and -10 codes.Separate control groups were created for patients without anorexia nervosa and for patients without bulimia nervosa.Both the experimental and control groups were generated from the general population available within the PearlDiver database.These control groups were propensity score matched by age, sex, Charlson Comorbidity Index, record dates, and geographical region.The anorexia cohort was thus compared to a matched cohort of patients without anorexia nervosa, and the bulimia cohort was compared to a matched cohort of patients without bulimia nervosa.All groups were composed of members of the general population, with no specific selection for athletic participation or level of physical activity, as this information is not encoded in the PearlDiver database.

Variables and outcome measures
For patients included in this study, instances of lower extremity soft tissue injuries were identified through ICD-9 and ICD-10 codes within the time period of 2010-2020.Injuries investigated included meniscus tear, medial collateral ligament sprain, lateral collateral ligament sprain, anterior cruciate ligament tear, posterior cruciate ligament tear, any cruciate ligament tear, calcaneofibular ligament sprain, deltoid ligament sprain, tibiofibular ligament sprain, unspecified ankle ligament sprain, and any ankle ligament sprain.Instances of soft tissue orthopedic surgery procedures in these patients were identified through Current Procedural Terminology (CPT) codes within the time period of 2010-2020.Procedures investigated included anterior cruciate ligament reconstruction, posterior cruciate ligament reconstruction, partial meniscectomy, any meniscus surgery, collateral ligament repair, and any ankle ligament repair.ICD and CPT codes used to define these outcomes of interest are detailed in Supplemental Table S1.

Statistical analysis
The prevalence of lower extremity soft tissue injury and orthopedic surgery among experimental and control cohorts were extracted and tabulated.The relative prevalence was then analyzed using univariate chi-squared analysis.A p-value <0.05 was used to determine the threshold of significance between experimental and control groups.A risk ratio and corresponding 95% confidence interval were calculated for each outcome of interest.

Demographics, description of study population
The anorexia nervosa cohort and its matched control group contained a total of 2,475 patients each.Patients in the anorexia nervosa cohort and corresponding matched control group had an equal distribution of ages (both cohorts: mean 36.9 ± 18.0 years), Charlson Comorbidity Index scores (both cohorts: mean 0.35 ± 0.73), and sexes (both cohorts: 90.9% female, 9.1% male).A full presentation of patient characteristics within the matched cohorts is detailed in Table 1.
The bulimia nervosa cohort and its matched control group contained a total of 2,378 patients each.Patients in the bulimia nervosa cohort and the corresponding matched control group had an equal distribution of ages (bulimia cohort mean: 36.2 ± 14.6 years, control cohort mean: 36.3 ± 14.6 years), Charlson Comorbidity Index scores (both cohorts: mean 0.29 ± 0.61), and sexes (both cohorts: 93.7% female, 6.3% male).The full presentation of patient characteristics within the matched cohorts is detailed in Table 1.

Background and rationale
The link between eating disorders and fracture risk has been well studied.This study expands our current knowledge to include soft tissue injuries and surgeries.Specifically, anorexia nervosa and bulimia nervosa are associated with an increased relative prevalence of lower extremity soft tissue injuries and surgeries.As the prevalence of eating disorders rises [4][5][6], it follows that there will be an increase in the number of orthopedic patients presenting with soft tissue injuries and surgeries.Consequently, surgeon awareness of these findings is critically important in providing appropriate care to patients with existing anorexia or bulimia diagnoses as well as education of patients at risk for eating disorder development.

Relative prevalence of soft tissue injuries
This study found that the prevalence of lower extremity soft tissue injuries increased in patients with anorexia nervosa or bulimia nervosa.Prior literature has proposed numerous mechanisms behind the increased fracture risk in eating disorder patients, and these studies have yielded some important insights that inform the interpretation of the results of this study.In particular, the vast majority of literature has centered around fracture risk being the result of the reduction of bone mineral density in patients with eating disorders, predominantly anorexia nervosa [13,14,20].However, some studies have postulated additional contributors to fracture risk, with one in particular demonstrating that lower leg lean tissue mass is correlated with stress fracture risk [15].Furthermore, as eating disorders are associated with a high degree of perfectionism and drive for thinness, it follows that athletes with eating disorders may place themselves at increased risk for injury due to compulsive overexercise [35,36].Among these three postulated drivers of injury risk, the latter two may also contribute to the increased prevalence of soft tissue injuries observed in this study.
Another potential driver of soft tissue injuries in patients with anorexia nervosa and bulimia nervosa is the impact of malnutrition on muscle function.Numerous studies have shown that patients with anorexia have reduced muscle size and energy expenditure relative to controls [30].These reductions in muscle size and function may reduce their ability to resist deforming forces, thus leaving patients vulnerable to soft tissue injuries.Furthermore, both anorexia and bulimia can result in severe electrolyte disturbances, as well as the dramatic changes in body composition seen in anorexia.One study in patients with anorexia nervosa demonstrated that muscle function upon electrical nerve stimulation was significantly reduced, with an increase in the force of contraction, slowing of the maximal relaxation rate, and an increase in muscle fatigability seen.Interestingly, with 4 weeks of refeeding and restoration of electrolyte levels, muscle function normalized [31].Consequently, this may point to a potential intervention to prevent soft tissue injuries, as refeeding and electrolyte restoration may normalize muscle function.

Relative prevalence of soft tissue surgeries
Our study also found that the prevalence of certain lower extremity soft tissue surgeries was increased in patients with anorexia nervosa or bulimia nervosa.It follows that with an increased injury prevalence would come an increased prevalence of surgeries.However, although there was an increased prevalence of meniscus surgeries that paralleled the increased prevalence of meniscus injuries in both cohorts, patients with anorexia or bulimia were not observed to have an increased prevalence of ankle surgeries despite having an increased injury prevalence, and patients with anorexia had an increased prevalence of anterior cruciate ligament (ACL) reconstruction despite not having an increased prevalence of ACL tears.Several reasons may be postulated for the disparities in injury rate and surgery rate.For example, surgical reconstruction is typically indicated for unstable ACL injuries [37], and, due to the effects of malnutrition on joint support postulated in this study, patients with eating disorders may be more likely to have more severe ACL injuries or multi-ligament injuries, and thus more likely to require surgery despite similar injury rates.Alternatively, as eating disorders are more prevalent in athletes, ACL surgery may be more likely in this population as surgery has been shown to have higher rates of return to preinjury level in sports [37].In ankle ligament injuries, however, conservative modalities make up the majority of treatments [38,39], thus athlete patients with eating would not be any more or less likely to undergo surgery.Future work should examine specific injury severity and decision-making regarding indication for surgery in patients with anorexia or bulimia.

Implications
There are several practical implications of our findings.First, it is important to consider how these drivers of injury risk impact care of orthopedic patients with anorexia nervosa or bulimia nervosa, particularly with regard to whether or not these risk factors are modifiable.A recent randomized control trial demonstrated that in exercising women with oligomenorrhea/amenorrhea who increased caloric intake for 12 months, there was no change in bone mineral density relative to controls despite increases in fat mass and body fat percentages [40].The results of this trial suggest that reduced bone mineral density as a contributor to stress fracture risk may be an unmodifiable risk factor, or that bone mineral density may take time in the span of years to increase, thus leaving individuals at elevated risk for stress fracture for extended periods of time.In the case of increased prevalence of soft tissue injury, we postulate that the mechanism of injury is based predominantly on muscular and nutritional deficiencies that may impact joint integrity, as well as eating disorder patients' tendency toward compulsive exercise.These factors, unlike bone mineral density, may be modifiable through behavioral changes such as resumption of proper nutritional intake and initiation of psychotherapy.Consequently, promotion of eating disorder recovery may prove to have an even greater benefit in patients' risk for soft tissue injuries than in risk for stress fractures.
Another practical implication of these results beyond physical outcomes is that promotion of musculoskeletal health in patients with anorexia or bulimia has implications for patients' wellbeing.For example, as eating disorders are associated with increased prevalence of injuries that result in time away from sport [29], surgeons may need to consider a patient's eating disorder diagnosis, or even suspicion of disordered eating, in determining clearance for return-to-play [11,41].Additionally, the results of this study demonstrated that patients with eating disorders had a higher relative prevalence of various knee and ankle surgeries.Undergoing these surgeries is associated with significant financial costs, as well as the risk of future surgery and disability due to osteoarthritis [42,43].Orthopedic surgeons may thus be in a crucial position to intervene in patients presenting with anorexia or bulimia and thus prevent incomplete recovery, future time away from sport, or financial hardships for these patients.
Several strategies may exist for orthopedic surgeons to implement understanding of the association of anorexia and bulimia with soft tissue injuries and surgeries in practice.Notably, in recent years addressing comorbid eating disorders has been incorporated into the management of orthopedic patients by the implementation by numerous institutions of women's sports medicine programs [44,45].However, these programs comprise a minority of orthopedic surgeons and are not designed to address male patients with eating disorders.Therefore, patients may benefit from an increased consideration of the symptoms of anorexia or bulimia by orthopedic surgeons beyond those involved in these programs.Recent studies have proposed mechanisms for increasing this consideration such as obtaining buy-in from leadership, integrating screening questions into visits, didactic training, designating 'mental health champions' with specific training in recognizing certain disorders, and addressing stigma [46][47][48].Patients with eating disorders are often hesitant to seek care [49], thus many presenting to an orthopedic surgeon may lack a diagnosis in their charts.As such, implementation of the aforementioned strategies may aid in not only identifying patients with anorexia or bulimia but also in promoting holistic care of these patients and hopefully reducing risk for future soft tissue injuries and surgeries.

Limitations
Although the results of this study provide a valuable addition to the literature regarding the care of orthopedic patients with anorexia nervosa and bulimia nervosa, we recognize that limitations exist.First, due to the retrospective nature of this study, it may be difficult to establish causation, rather than association.This study attempted to mitigate this limitation through the use of the PearlDiver database, which allows for longitudinal monitoring of individual patient outcomes, rather than viewing data crosssectionally.Second, although the control cohort was generated with matching for age, sex, Charlson Comorbidity Index, record dates, and geographical region, we recognize that there could be other mediating factors such as BMI, participation in sports, race/ethnicity, and socioeconomic status, which we were unable to control for.For example, patients with anorexia may be more likely to participate in sports that inherently predispose them to the injuries investigated in this study, and patients with bulimia may be more likely to use compulsive exercise as a compensatory behavior after binge eating, thus increasing risk for exercise-related injuries.However, we note that the fact that the prevalence of eating disorders in athletes is elevated does not necessarily mean that the inverse is true, that a greater percentage of the population of eating disorder patients are athletes than in the general population.For example, the population of eating disorder patients is also composed of other high-prevalence groups such as non-athletic performers such as actors and fashion models [50,51], patients with autism [52], sexual abuse survivors [53,54], transgender and gender diverse individuals [55,56], and veterans [57].Consequently, the patients in our cohort may have been just as likely to belong to one of these groups as to have been athletes, thus potentially reducing the impact of the limitation that we could not control for athlete status.A third limitation is that our cohorts for several injury and surgery types were relatively small.Therefore, it is possible that the relative prevalence for these injuries and surgeries may have been over-or understated and thus should be interpreted with caution.Similarly, there were <11 patients from the Western region in each of our cohorts, which may limit the generalizability of our results.Finally, eating disorders are frequently underreported or underdiagnosed, particularly in male and minority populations [58][59][60][61].Therefore, some patients within our control cohort may have had eating disorders not represented in their medical records.Similarly, the data likely do not capture patients with disordered eating or subclinical eating disorders, which may also be associated with soft tissue injuries.However, the effect of this inclusion in the control cohort of patients with eating disorders or subclinical eating disorders would potentially be to decrease the relative risk of injury or surgery, thus limiting the bias introduced by inclusion of these patients in our control.

Conclusions
In summary, diagnosis of anorexia nervosa or bulimia nervosa is associated with an increased prevalence of certain lower extremity soft tissue injuries and surgeries.Orthopedic surgeons should be aware of the effects these disorders may have on soft tissue injury and surgery rates.Furthermore, patients presenting to orthopedic surgery clinics should be informed of the association of their eating disorder diagnoses with injury and surgery prevalence and provided with information on resources promoting recovery to help prevent further injury.

Table 1 .
Descriptive statistics of anorexia nervosa and bulimia nervosa matched cohort demographics.

Table 2 .
Prevalence and relative risk of soft tissue injuries in anorexia nervosa and bulimia nervosa cohorts vs. controls over 2010-2020.
Note: *Cohorts were matched by age, gender, Charlson Comorbidity Index, region, and dates of care, and significance was determined by chi-squared test.P-values were adjusted with Yates' continuity correction.Bold implies p<0.05.

Table 3 .
Prevalence and relative risk of soft tissue orthopedic surgeries in anorexia nervosa and bulimia nervosa cohorts vs. controls over 2010-2020.Cohorts were matched by age, gender, Charlson Comorbidity Index, and dates of care, and significance was determined by chi-squared test.P-values were adjusted with Yates' continuity correction.Bold implies p<0.05.