Clinical characteristics in patients with cervicogenic dizziness: A systematic review

Abstract Background and aims Cervicogenic dizziness (CD) is a clinical syndrome of dizziness associated with neck dysfunction. CD represents a considerable diagnostic challenge since dizziness and neck pain are common symptoms with complex and multifactorial etiologies. Both research and clinical work on CD is limited by the lack of accepted diagnostic criteria. The aim of this study was to review clinical studies on CD and to assess current evidence regarding the clinical characteristics of this syndrome. Methods A comprehensive PubMed and MEDLINE search was conducted from the date of inception of the database, with the last search conducted in September 2018. Included studies had to contain operable diagnostic criteria as well as a comparison between patients considered to have CD and a clinical comparison group. Data extracted were clinical outcomes, diagnostic criteria, age, sex, and sample size. Studies were assessed for methodological quality using the Crowe Critical Appraisal Tool. Results Out of 2161 screened studies, eight studies comprising 225 patients met the inclusion criteria. Studies were of low to acceptable methodological quality. The most frequent and consistent clinical characteristic in patients classified as having CD, compared with other populations, was reduced posturographic stability. The most consistent diagnostic criteria were based on the concurrence of neck pain with dizziness after exclusion of other possible reasons for dizziness. Conclusion There are few studies examining clinical characteristics in patients with cervicogenic dizziness. Altered posturography appeared to be the only consistent characteristic used when distinguishing CD from other populations. Diagnostic criteria currently used in research are likely to have low specificity, since they rest on the exclusion of other causes rather than on positive distinctive features. More studies are needed to better understand the clinical interrelations between dizziness and neck pain.


| INTRODUCTION
The clinical diagnosis of cervicogenic dizziness (CD) is commonly reserved for patients presenting with dizziness associated with neck dysfunction after all other potential causes for the dizziness have been excluded. 1 However, the usefulness of this diagnostic approach in a clinical setting is limited for several reasons.
Dizziness is a common symptom that may arise from a great number of disorders. 2,3 In approaching the dizzy patient, it is essential to narrow down this number by assessing symptoms, their time course, and possible triggers. The word "dizziness," in itself, is insufficient to qualify as a diagnostic criterion. Typical clinical symptoms of CD are suggested to consist of disorientation, lightheadedness, or disequilibrium accompanied by cervical pain, limited range of motion, and reduced balance. 4,5 In addition, a close temporal relationship between the dizziness and neck symptoms is considered important by some authors (Wrisley et al 2000). An ex juvantibus confirmation of the diagnosis-based on the resolution of dizziness after treatment of the neck disorderhas been proposed. 1 However, clinical studies documenting the vestibular or extra-vestibular symptoms, whether they be vertiginous or not, whether acute, episodic, or chronic, or triggered by specific activities or events, are needed. CD has several proposed causes, such as vascular or neurovascular. 6 However, the most common theory is considering CD to be a disorder of neck proprioception. 1,7 Furman and Cass 7 defined it as a "nonspecific sensation of altered orientation in space and disequilibrium originating from abnormal afferent activity from the neck." Because of high demands of both stability and mobility, the cervical spine has a well-developed proprioceptive system. [8][9][10] Thus, the functional status of the neck should be examined, and the use of neck pain as a diagnostic marker of CD may, therefore, be inadequate. Thus, there is a need for clinical studies documenting neck function in patients with CD.
To date, there is no consensus on diagnostic criteria for CD. Several reviews have been published on the topic, but these have mainly focused on the theoretical basis for the diagnosis, eg, the abundance of muscle spindles in the deep cervical muscles, 10 the close integration between cervical and vestibular afferents in the brain stem and cerebellum, 11 and experimental studies on the effect of selective neck lesions or injection on balance and dizziness. [12][13][14][15] To the authors' knowledge, no systematic review exists of clinical studies on CD and how these patients differ from other relevant patient populations such as those with other diagnosis of dizziness, patients with neck pain, or even healthy controls. Identifying studies examining how CD patients differ from other populations would contribute to better understand the condition and guide future research.
The aim of this paper is to review clinical studies on CD and to assess current evidence regarding the clinical characteristics of this syndrome. A secondary aim was to examine and compare the diagnostic criteria that were used in the included studies.

| Eligibility criteria
This review was restricted to published, peer-reviewed original studies. Unpublished studies, case reports, editorials, reviews, and conference abstracts were not included. The search was restricted to articles written in English. We included original studies on patients with CD because of allegedly altered neck proprioception, comparing their clinical characteristics to those of other populations. Thus, for inclusion, the study had to contain a reference group, either with another diagnosis or healthy controls, for comparison. To assure higher comparability between studies, included studies had to state whether or not other causes of dizziness had been ruled out. This included other causes of alleged CD such as neurovascular or vascular disorders. In addition, the diagnostic process or criteria had to be accounted for.
Studies were excluded if the study population (CD) was composed of patients suffering from other confirmed diseases that could explain their symptoms. For readability and consistency, this review uses the term CD, although some of the included papers have used slightly different names for the same condition (Table 1).

| Study selection
All titles and abstracts were screened by the two reviewers after duplicates were discarded and irrelevant citations were removed. Full text versions of eligible articles were evaluated by the two reviewers to determine inclusion. Any disagreements were resolved through discussion among reviewers. The process was facilitated by the use of the Rayyan systematic review web application, 16

| Data extraction process
The following data were extracted, compared, and compiled in a spreadsheet by both reviewers: population (age, sex, and sample size), study design, diagnostic criteria, and clinical findings compared with other diagnosis/healthy controls. The two reviewers compared the entered data and corrected missing entries.

| Assessment of methodological quality
Because of the heterogeneous nature of the studies with regard to design and outcome measures, quality of data, and study design, a meta-analysis was not appropriate for this review. Thus, a quality analysis of the included studies was performed. The methodological quality of the studies was assessed using the Crowe Critical Appraisal Tool version 1.4 (CCAT), which allows for a variety of research designs to be evaluated using the same tool. 18 This tool consists of nine categories. The first eight categories have a score range from 0 to 5. The ninth category states the total sum from the previous eight categories, which can range from 0 to 40, where a higher score indicates higher quality.

| Ethical Approval
Ethical approval was not required for this systematic review.

| Search
The search resulted in 2161 articles, after removing duplicates.
After screening titles and abstracts for irrelevant citations, we identified 59 articles, which were assessed in full text. No additional articles were found when screening reference lists. Fifty-one studies did not meet the inclusion criteria and were excluded from this review. See Appendix S2 for a list of excluded studies with reasons for exclusion. A total of eight studies met the inclusion criteria.
The selection process is shown in Figure 1. The eight included studies comprised four cross-sectional studies, [19][20][21][22] one prospective study, 23 two case-control studies, 24,25 and one single-subject design study. 26 The included studies comprised a total of 225 patients classified as CD, with group sizes ranging from n = 14 to 86.
Patients were compared with healthy controls (n = 140) in five studies, 19,[21][22][23]26 to patients with BBPV (n = 25) in one study, 24 to patients with general dizziness (n = 86) in one study, 25 to patients with vestibular neuritis (n = 18) in one study, 21 and to patients with only neck pain (n = 40) in two studies. 19,22 Most studies included more women (n = 136) than men (n = 89), with the percentage of women ranging from 42 % to 87 %. The age of the CD patients ranged from 36 to 66 years. The included studies, with methodological quality assessment, are shown in Table 1.

| Posturography
A total of five studies included posturography. One of the studies found that the posturographic response to vibratory stimulation of the calf muscles could distinguish patients with vertigo of suspected cervical origin from patients with vestibular neuronitis and healthy controls. 21 Two of the studies found that patients with CD had reduced postural control compared with both patients with only neck pain and healthy controls. 19 The last two studies found reduced postural control in CD patients compared with healthy controls. 20,23

| Cervical proprioception measured by relocation tests
Two studies examined cervical proprioception using relocation tests.

| Cervical range of motion
Cervical range of motion (CROM) was examined in two studies, with different measurements methods. 19 Yahia et al. 22 found that patients with chronic neck pain and vertigo had significantly lower CROM (measured in centimeters from chin to sternum, chin to acromion, and earlobe to acromion) compared with both patients with only chronic neck pain and healthy controls. Alund et al. 19 found no difference in CROM (measured with a three-dimensional electrogonimetric equipment) between patients with suspected CD, neck pain, and healthy controls.

| Symptom duration
Two of the studies reported duration of dizziness. In one of the studies, the patients with CD had longer duration of dizziness (81 months) compared with patients with general dizziness (23 months). 25 In the other, the patients with CD exhibited shorter dizziness duration (30 months) compared with patients with BPPV (38 months). 24

| Neck pain
Neck pain was examined in three studies. L'Heureux-Lebeau et al. 24

| Psychometric measures
L'Heureux-Lebeau et al. 24 found no difference in anxiety or dizziness handicap between patient with CD and those with BPPV, using the Dizziness Handicap Inventory and State-Trait Anxiety Inventory.
Grande-Alonso et al. 20 found that patients with CD had higher fear of movement and higher anxiety and depression levels than asymptomatic individuals, as measured by the Tampa Scale for Kinesophobia and Hospital Anxiety and Depression Scale.

| Dizziness characteristics and triggers
Only one study examined differences in dizziness characteristics between patients with CD and other dizziness diagnoses. The study found that patients with CD were more likely to have a sensation of drunkenness/lightheadedness 24 compared with patients with BPPV.
Patients with BPPV were more likely to experience rotatory vertigo.
The CD group was more likely to report cervical movement as a precipitating factor. There were no differences in self-reported imbalance, dizziness, lightheadedness, floating sensation, sway sensation, nausea, falls, or dizziness frequency between the two groups. Reid et al. 25

| Headache
One of the included studies 22 found that patients with chronic neck pain and vertigo had more neck-related headaches compared with patients with only chronic neck pain.

| Coexistence of dizziness and neck pain
All but one 26 of the included studies had the coexistence of neck pain and dizziness as an explicit diagnostic criterion. In Heikkila et al., 26 neck pain was implicated in the criterion "dizziness or vertigo of suspected cervical origin."

| Vestibular symptoms, triggers, and aggravating factors
Most of the included studies did not specify particular dizziness symptoms as criteria for classifying patients as CD. However, one study 25 included dizziness "described as unsteadiness triggered by neck movement" as a criterion. Another study included dizziness "associated with pain, movement rigidity, or certain neck positions" as a criterion. 20

| Timing and duration of neck symptoms and dizziness
Four of the included studies specified duration of symptoms in the diagnostic criteria. One study reported that the patients had to have "recent onset" of and simultaneous complaint of dizziness or vertigo. 21 Another reported that the duration of both neck pain and dizziness had to be longer than 3 months. 20 Yahia et al. 22 used chronic neck pain of more than 3-month duration as a criterion. Alund et al. 19 chose neck pain and stiffness for more than one year as a criterion.
The criteria for dizziness were only reported as "long-lasting." Finally, one study added that if the neck pain had a traumatic origin, there needed to be a temporal proximity between the onset of dizziness and the neck injury. 24

| Neck examination
Two studies included decreased neck mobility in the diagnostic criteria. 19 Reid et al. 25 reported stiff and/or painful neck as one of their criteria, whereas Alund et al. 19 mentioned "localized neck pain and stiffness." Reid et al. 25 additionally required "palpable upper cervical spine dysfunction" assessed by an experienced physical therapist.

| Other causes excluded
All studies reported exclusion of causes of dizziness/vertigo, such as vestibular and central. The studies described in detail the method and examination used for ruling out patients with other causes of dizziness or vertigo, except for one. 20 However, this study noted that presence of an otorhinolaryngological diagnosis of central or peripheral vertigo would exclude the patient from their study.

| Methodical quality of the studies
The studies were given CCAT scores ranging from 14 to 28, indicating low to acceptable methodical quality. Common limitations in the included studies were insufficient information on sampling methods, insufficient sample size justification, insufficient information on ethical matters, and limitations related to statistical analysis.

| DISCUSSION
This review identified eight original studies comparing patients with CD with groups of patients either suffering from other established and well-defined conditions or healthy controls. Based on CCAT scores, the studies were of low to acceptable methodological quality.
Pooling of the results was not possible since outcomes varied. Nevertheless, the studies shed some light on current opinions on CD.

| Clinical findings
Although the International Classification of Vestibular Disorders distinguishes between vertigo and dizziness, 27 and some consider it unlikely that disorders of neck proprioception should be associated with illusory perceptions of self-motion such as spinning vertigo, 28 only one study in this review 25 required the a priori exclusion of patients with vertigo, stressing that dizziness should be described as found that 32% of patients with CD reported a rotatory sensation compared with 76% in a group with BPPV. In this study, most patients reported a sensation of "drunkenness" (92%) or imbalance (76%).
Admittedly, one should not rely solely on the description of vestibular symptoms in making a diagnosis, since patients have difficulties reporting vestibular symptoms in a consistent way. 29

| Diagnostic criteria for cervicogenic dizziness
The diagnostic criteria used in the reviewed studies were predefined by the authors, and because of the lack of a diagnostic "gold standard," their validity cannot be determined. The criteria of CD was, in most studies, based on the patient simultaneously reporting neck pain and dizziness as well as the exclusion of other neurological or neuro-otological disorders. The distinction between vertigo and dizziness was not considered essential for the diagnosis in most of the reviewed studies. One study specified that the dizziness should be described as "unsteadiness", 25 while another required vertigo defined as an "erroneous impression of the movement of objects relative to the subject or the movement of the subject relative to his/her environment." Neck stiffness or rigidity was not usually required for the diagnosis but mentioned in the inclusion criteria of three studies. 19,25 The same was the case with localized tenderness in the neck, which was mentioned in two studies. 19,25 Positive objective signs were usually not considered necessary, except for one study 25  It has been argued that the diagnosis of CD may be mainly of exclusive academic interest, since the treatment is often the same as for patients with cervical pain syndrome. 28 However, a correct diagnosis will always be clinically meaningful in guiding the treatment and in reassuring the patient that an explanation for their distressing symptoms has been found. Lastly, a conclusive diagnosis could save both the patients and the health care system from the consequences of unnecessary diagnostic and therapeutic procedures.

| LIMITATIONS
The review was limited to studies reported in English. Because of the low number, varying outcomes, and the low to moderate methodolog- Diagnostic criteria differed between studies and were mostly based on the coexistence of neck pain with dizziness and the exclusion of other neurological and neuro-otological causes. Thus, the sensitivity and specificity of the criteria are likely to be low. As this review revealed significant differences in methodical and experimental approaches, this should be considered when designing future studies, making comparison between studies more feasible.