The Prevalence of Pain in People With Chronic Ankle Instability: A Systematic Review

All studies were identified through a search of electronic databases, including AMED, CINAHL, EMBASE, MEDLINE, PubMed, Scopus, SPORTDiscus, and Web of Science from inception until March 2017. No language restriction was imposed. Key terms used in the search strategy were based on broad terms and related synonyms targeting 3 categories: ankle-sprain injury, joint instability, and ankle pain (Supplemental Table 1, available online at http://dx.doi.org/10.4085/1062-6050-531-17.S1).

This study protocol was developed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines²⁵  and was registered on PROSPERO (registration number CRD42015024285). Studies of CAI were included if they met the following inclusion criteria: (1) the studies were peer reviewed (observational or randomized controlled trials); (2) participants were of any age; (3) participants had a history of LASs and ongoing impairments, such as pain, giving way, recurrent sprains, or feelings of instability; (4) patients were followed for more than 3 months after the initial LAS; and (5) patients had no associated fractures. Studies were excluded if they were incomplete (eg, abstracts), case studies, or reviews or included only surgical interventions. Studies of participants with medial ankle injuries or syndesmosis injury alone were also excluded.

Articles identified by the search strategy were screened independently by 2 authors (S.A. and C.H. or F.P.) for the inclusion criteria using the title and abstract and then the full-text papers. Any discrepancies regarding inclusion were resolved by consensus. Data were extracted independently by 2 investigators (S.A. and C.H., M.M., or F.P.). A third author adjudicated when the first author (S.A.) and any of the other 3 investigators disagreed. Data extraction involved the following items: author and year, study type, sample size, participant demographics, target population, inclusion criteria, outcome measures, and study results. Outcomes were (1) pain ratings using measures such as a visual analog scale or physical examination and (2) other residual impairments, such as feelings of weakness, giving way, reduced balance, or deficits in functional performance.

The methodologic quality of the included studies was assessed independently by 2 reviewers (S.A. and C.H., M.M., or F.P.) using a modified version of the Downs and Black assessment tool.²⁶  The modified tool consisted of 25 items that covered the following quality characteristics: reporting, external validity, internal validity (ie, bias, confounding), and power. One of 4 answers could be selected when assessing each item: yes (score = 1), no (score = 0), unable to determine (score = 0), and not applicable (score = NA). Each study had a different total score, as NA answers were not included; therefore, the percentage of the total score was reported.

For the data synthesis, we calculated the total number of participants with CAI and then calculated the proportion of participants who still reported pain. Included studies were divided according to their pain-assessment method at the ankle (self-report or manual test) and ordered according to their quality-assessment results (Table). Data were presented as the percentage of pain in participants with CAI over different follow-up periods from the initial time of injury.

The initial search identified 5907 papers. After duplicates were removed, the titles and abstracts of 4140 papers were screened for potential eligibility. After the initial screening, we identified 154 articles as potentially eligible and sought full texts. After full-text screening, 14 full-text papers were included in the review (Figure 1).

Included studies comprised 9 prospective cohort studies,^{2,8,28–30,32–35}  2 cross-sectional studies,^10,36  1 self-administered study,²⁷  1 observational study,³¹  and 1 retrospective study.¹¹  Follow-up periods varied from 3 to 24 months after an LAS in some studies,^{2,28,29,32–35}  whereas they were more than 6 years in other studies^8,10,11  (Supplemental Table 2). The age of participants varied among studies, with most participants classified as young. The participants' mean age ranged from 20 to 35 years in 6 studies,^{2,28,30,33,35,36}  from 16 to 65 years in 7 studies,^{8,10,11,27,31,32,34}  and from 8 to 18 years in 1 study.²⁹  Participants were recruited from different settings, including hospitals or primary care practices or both,^{8,11,27–29,31,34,35}  military populations,^2,32,33  and metropolitan areas.^10,30,36 

Authors of 11 studies used a self-reported measure of pain outcome,^{8,10,11,27–32,34,35}  whereas researchers in 2 studies used either a pain-provoking physical examination procedure, such as a varus stress test and palpation,²  or ankle end range of motion.³⁶  Gerber et al³³  used both a self-report measure and physical examination for pain assessment over 6 months. In investigations using a self-report measure, pain was measured by visual analog scale,^28,31,33,34  survey questions,^10,27  telephone call,^8,29,32,35  or pain questionnaires.^8,11 

In 9 studies using self-reported pain measures, the percentage of participants with CAI who had pain ranged from 18% to 79%, with an average of 58% (Figure 2).^{8,10,11,27,29,31–33,35}  None of these authors used pain as the primary outcome. No researchers investigated the relationship between the presence of pain in participants with CAI and other residual impairments of CAI.

Most included studies were rated as high quality using the modified Downs and Black quality-assessment tool. The percentages of scores ranged from 54% to 100%, with a mean of 80.3%. Eleven of 14 studies scored from 71% to 100%, with a mean of 84.8% in total assessment points,^{2,8,10,27–33,36}  whereas 3 studies scored from 54% to 69%, with a mean of 64% (Supplemental Table 3).^11,34,35  In most studies, some criteria were not applicable, and they were not included in the total score. Low-quality characteristics identified in some studies included a lack of blinding of the study participants to the intervention, no randomized intervention, not reporting the number of participants lost to follow-up, and insufficient study power.²⁶ 

In this systematic review, all included studies revealed pain as one of the residual impairments that was present for a prolonged time in participants with CAI. Generally, 55.4% of participants with CAI reported pain. Studies in which researchers used self-report measures of pain demonstrated a higher percentage of participants reporting pain (average = 58%) than those in which researchers used a physical provocation examination to investigate the presence of pain (average = 49%). No investigators examined the association between pain and other residual impairments among participants with CAI.

Pain was assessed by self-report and physical provocation examination. Among participants with CAI who had pain assessed by self-report at a single time, the 4 (out of 7) highest-quality studies showed that more than 66% of participants with CAI reported pain for a prolonged time (6 months to 7 years) after the initial injury.^8,10,27,29  This prevalence of pain was consistent across time. In contrast, Gerber et al³³  demonstrated that approximately 44% of participants with CAI had pain at 6 months after the initial injury. They focused on a young military population that may have been reluctant to admit having pain 6 months after the injury.³³  They stated that one of the reasons that soldiers returned to activity before completing their rehabilitation programs was concern about losing their positions after a prolonged absence. Much less pain prevalence was also shown in the 2 prospective studies.^11,32  In 1 study, the percentage of pain in participants with CAI declined from 50% at 3 months to 0% at 12 months.²⁸  The decline was probably due to the high dropout rate: from 21 to 2 participants. Verhagen et al¹¹  found a low prevalence of pain (28%) at 9 months after injury, which decreased by 6.5 years (18%; χ² = 4.896, P = .03). The lower pain level may have been due to the standardized rehabilitation program participants followed in the first 6 weeks after the injury, which included a walking cast, partial immobilization with tape, and physiotherapy.

Studies that assessed pain by physical provocation examination^2,33,36  included palpation,³³  varus stress test,²  and palpation or end–range-of-motion tests.³⁶  In the varus stress test, approximately 75% of participants with CAI experienced lateral ankle pain while in a plantar-flexed position, whereas 25% had pain in the neutral position.²  Kwon et al²  suggested that pain during the varus stress test may be correlated with a complete tear of the anterior talofibular ligament. However, they stated that ascertaining ligament status using magnetic resonance imaging could be biased. Whereas the test in the plantar-flexed position yielded a high percentage of participants reporting pain, the stresses applied in this test position would not be expected to occur during activities of daily living.

The effect of age on self-reported pain was unclear, as most studies included heterogeneous groups of participants with different age ranges.^{8,10,11,27,28,30–32,34,35}  The prevalence of pain in these studies ranged from 28.2%¹¹  to 79%³⁵  (Table). Only Gerber et al³³  focused on younger adults (age range, 17–24 years) and provided the percentage of CAI participants with pain (44%). Timm et al²⁹  evaluated children and adolescents (age range, 8–18 years) and found a pain prevalence in the upper range (69.1%) of the other studies.^{8,10,27,28,32,35}  No studies looked at older adults only. Whereas researchers³⁷  have suggested that older adults are more likely to be susceptible to the negative effects of pain, we could not find any work addressing the effect of age on reported pain among participants with CAI. Given that the effects of physiological, psychological, and social factors vary in younger and older adults and all these factors affect pain,^38–40  investigators should determine the effect of age on pain prevalence and severity among participants with CAI.

The frequency and severity of and activities that caused pain in participants with CAI were provided in some studies.^{8,10,31,34,35}  Comparing chronic ankle disorders, Hiller et al¹⁰  showed that 51.6% of participants with CAI reported occasional ankle pain; the severity of pain was mild in 33.9% and moderate in 29.0%. Whereas this was the only study in which researchers reported the frequency and severity of pain, definitions of these terms were not supplied. Activities that caused pain were walking (8%),³¹  running (22%),³¹  and movements after sports (25%).²⁴  Overall, most subjective nonstandardized questionnaires described CAI pain as being intermittent, mild, and only incited by vigorous activity.

We found no studies in which researchers investigated the effects of pain on the residual impairments of CAI. Given that pain is known to affect functional activity,^11,41  we could hypothesize that participants with CAI who have pain may show differences in various functional performance measures compared with participants who do not have pain. Hass et al⁴²  demonstrated that chronic peripheral joint injury (ie, CAI) negatively altered the supraspinal aspects of motor control. These changes strongly suggested that centrally mediated changes to motor control may be important contributors to the underlying neurophysiological mechanism of CAI.

Chronic pain was associated with sensitization of the neural network that promotes pain and changed circuits in the nociceptive pathways.¹⁷  In an animal model, these changes were evident in the representation areas of pain in the thalamus and cortex.⁴³  Patients with chronic pain were hyperactive to noxious stimuli, and hyperactivity increased with pain chronicity.⁴⁴  Researchers noted that the pain threshold and pain tolerance were lower among patients with chronic back pain^44,45  or knee osteoarthritis.⁴⁶  These results demonstrate that the changes in the central nervous system play a major role in chronic pain. Given that participants with CAI may have central alternations, investigating these changes associated with pain in participants with CAI could be an important topic for researchers to address.

A limitation of this systematic review was that the authors of the included studies used mostly subjective methods of measuring and describing pain rather than validated questionnaires. Many investigators assessed populations whose characteristics were slightly broader than the International Ankle Consortium selection criteria for individuals with CAI.¹  Our inclusion criterion of 3 months since the initial injury was less than the 6 months recommended by the International Ankle Consortium; however, the results did not change when a 6-month criterion was applied.

Researchers should be aware of the criteria associated with CAI to create clear descriptions of CAI and avoid heterogeneous groups. Future investigators should incorporate both subjective and objective pain measures, such as pain-threshold tests. The role of pain in the model and definition of CAI and the influence on long-term outcomes should also be explored in future studies.

We highlighted the importance of investigating the effects of pain in participants with CAI. Validated measures of pain severity, such as the Numeric Rating Scale, and provoking activities should continue to be monitored by clinicians, as pain appears to persist for a greater time than previously appreciated.

Ankle pain in participants with CAI is a common impairment that can potentially affect functional activity. The proportion of participants with CAI who still reported pain varied widely but was 50% to 79% in most studies. However, a large knowledge gap exists in describing ankle pain among participants with CAI and associated impairments, functional measures, and activities, which needs to be addressed.

Supplemental Tables. Key terms used in search strategy; included studies with more detailed information; results of modified Downs and Black's Quality Assessment Tool

Found at DOI: http://dx.doi.org/10.4085/1062-6050-531-17.S1