Abstract
Background
Neck pain is a frequent complaint among patients with migraine and seems to be correlated with the headache frequency. Neck pain is more common in patients with chronic migraine compared to episodic migraine. However, prevalence of neck pain in patients with migraine varies among studies.
Objective
To estimate the prevalence of neck pain in patients with migraine and non-headache controls in observational studies.
Methods
A systematic literature search on PubMed and Embase was conducted to identify studies reporting prevalence of neck pain in migraine patients. This review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Data was extracted by two independent investigators and results were pooled using random-effects meta-analysis. The protocol was registered with PROSPERO (CRD42021264898).
Results
The search identified 2490 citations of which 30 contained relevant original population based and clinic-based data. Among these, 24 studies provided data eligible for the analysis. The meta-analysis for clinic-based studies demonstrated that the pooled relative frequency of neck pain was 77.0% (95% CI: 69.0–86.4) in the migraine group and 23.2% (95% CI:18.6–28.5) in the non-headache control group. Neck pain was more frequent in patients with chronic migraine (87.0%, 95% CI: 77.0–93.0) compared to episodic migraine (77.0%, 95% CI: 69.0–84.0). Neck pain was 12 times more prevalent in migraine patients compared to non-headache controls and two times more prevalent in patients with chronic migraine compared to episodic migraine. The calculated heterogeneity (I2 values) ranged from 61.3% to 72.0%.
Conclusion
Neck pain is a frequent complaint among patients with migraine. The heterogeneity among the studies emphasize important aspects to consider in future research of neck pain in migraine to improve our understanding of the driving mechanisms of neck pain in a major group of migraine patients.
Introduction
Neck pain has been reported to be highly prevalent in migraine patients (1–3). Whether a causal link exists between neck pain and migraine, or if it is just a secondary phenomenon, is not fully clarified (4–6). Nociceptive input from neck muscles and dura mater converge on second-order neuron in the trigeminocervical complex (7–9). Due to these anatomical connections, activation of cervical nociceptive afferents may be involved in the induction and maintenance of migraine headache (10,11).
Studies have shown that the prevalence of neck pain correlates with headache frequency (12–14) and that neck pain is suggested as a coexistent feature of migraine (12,13). Neck pain is more prevalent in patients with chronic migraine compared to episodic migraine (15,16), and neck-related disability appears to be associated with migraine chronification (17). In addition, the reported timing of neck pain onset in relation to the migraine attack phases varies between studies (2,18).
Elucidating these aspects of the neck pain prevalence may provide clues to its potential pathophysiological role in migraine. Consequently, we conducted a systematic review and meta-analysis of the prevalence of neck pain in migraine patients and non-headache controls. In addition, we investigated potential differences in prevalence between patients with episodic and chronic migraine. Finally, we explored the association of neck pain to the different phases of a migraine attack.
Methods
Data sources
We conducted a systematic review according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guideline (19). Furthermore, this systematic review was registered with PROSPERO, the International Prospective Register of Systematic Reviews (registration number CRD42021264898).
We searched PubMed and Embase for articles concerning neck pain and migraine. The search was performed on 1 July 2021 with the following search string:
The search was performed from database inception until the date of the database search. Furthermore, search strategy was supplemented by a review of the reference lists of relevant primary articles to identify any studies missed in the initial search.
Study selection
Two investigators (HMA and ZA) independently performed the title and abstract screening, along with full text screening of relevant articles. Reasons for exclusion of any articles were recorded and summarized. The full eligibility criteria are presented in Table 1.
Inclusion and Exclusion Criteria.
Data extraction and quality assessment
Two investigators (HMA and ZA) extracted data independently. Following variables were extracted: study design, age, sex, assessment method of migraine, migraine subtype, assessment method of neck pain, number of patients with migraine, number of non-headache controls and timing of neck pain (before, during or after an attack). Inclusion of a study was decided by consensus between the two investigators (HMA and ZA). If any event of disagreement in the data extraction between the two investigators (HMA and ZA) occurred, one of two independent senior investigators (SY and HWS) were consulted to resolve the disagreement.
The risk of bias of the studies was assessed using the Joanna Briggs Institute (JBI) Critical Appraisal Tool for prevalence studies (20). Two independent investigators (HMA and ZA) assessed each study. Any disagreement between the two reviewers was resolved by consensus by a third author (SY). The key aspects of the JBI Critical Appraisal Tools for prevalence studies are the following: I) addressing the target population; II) sampled cases in an appropriate way; III) adequacy of sample size; IV) methods; V) data analysis; VI) correct method used for the identification of the condition; VII) reliable way to measure the condition; VIII) appropriate statistical analysis; IX) adequacy of response rate. Each question was assigned one score with the possible range of scores being 0–9.
A quality assessment of “High”, “Moderate”, or “Low” score was rated for each study giving ≥7 scores, 4–6 scores, and ≤4 scores, respectively.
Definition of neck pain
The definition of neck pain varied among studies. Therefore, we included studies reporting neck pain as a coexistent and associated complaint to migraine.
Data synthesis and analysis
A descriptive meta-analysis of eligible clinic-based studies was performed. The initial outcome of interest was the relative frequency of neck pain in migraine patients and non-headache controls. Furthermore, to analyse the likelihood of neck pain in migraine patients compared to non-headache controls and the likelihood of neck pain in chronic migraine patients compared to episodic migraine patients. The pooled relative frequencies of neck pain were estimated by pooling study specific estimates. Meta-analysis with random-effects were performed to investigate between-study heterogeneity. Relative frequencies were calculated for subgroups consisting of three or more study populations. The I2 statistic were used to assess between-study heterogeneity. A funnel plot of standard error against proportion effect size and quantitatively by Egger’s test (with significance defined as α ≤ 0.05) of studies was used to assess the visual inception of publication bias. Analyses were performed with the Comprehensive Meta-Analysis statistical software (version 3.0).
Results
The database search resulted in a total of 2485 hits (Figure 1). An additional five articles were identified through a manual search of the reference lists of relevant primary articles and reviews. A total of 2490 articles were screened by title and abstract. Investigator agreement for studies selected by title and abstract screening was good (Kappa = 0.85) (21). A total of 96 articles were selected for retrieval and further consideration. Thirty articles (published between 1994 and 2021) reported prevalence of neck pain in migraine patients according to the eligibility criteria and were included in the meta-analysis. Of these, 24 studies were clinic-based studies, while two were populations-based studies. Results extracted from the clinic-based and population-based studies are reported separately. The 24 clinic-based studies provided data for the quantitative analysis. Eight of the 24 clinic-based studies included healthy participants as controls. These non-headache controls did not suffer from migraine, tension-type headache, or other pain conditions.
Clinic-based studies
Two clinic-based studies collected data retrospectively (22,23), while the remaining 22 studies acquired data prospectively. None of studies reported whether the migraine patients also experienced tension type headache (TTH). Eleven studies included only female participants (14,16,17,24–31). Six studies included both migraine patients with and without aura (2,18,31–34), while one study excluded patients with aura (28).
Neck pain in migraine patients
The 24 clinic-based studies comprised of 4352 migraine patients (Table 2) (2,14–18,22–39). Meta-analysis produced a pooled relative frequency of neck pain of 73.2% (95% confidence interval (CI): 66.7–78.9; I2 = 61.3%).
Estimated Relative Frequencies of Neck Pain in Patients with Migraine.
*These numbers were calculated based on data presented in the result sections of the included articles. ICHD = International Classification of Headache Disorders; F = Female; M = Male; Migraine = Not specified if migraine was episodic or chronic; EM = Episodic migraine; CM = Chronic migraine; NR = Not reported; I2 = Inconsistency.
Neck pain in migraine patients vs. non-headache controls
Eight studies included non-headache controls (16,18,26,27,29,30,35,36), yielding a total of 564 migraine patients and 254 non-headache controls (Supplementary Table 1) (Figure 4). The pooled relative frequency of neck pain was 77.0% (95% CI: 69.0–86.4; I2 = 71.9%) in migraine patients and 23.2% (95% CI:18.6–28.5; I2 = 0.0%) in non-headache controls (Figure 1). The likelihood analysis showed that those with migraine were ∼12 times more likely to experience neck pain than non-headache controls (odds ratio (OR): 11.5, 95% CI: 5.8–22.4; P = 0.000; I2 = 71.1%) (Figure 2).

Flow Diagram of Neck Pain in Migraine.
Neck pain in episodic vs. chronic migraine patients
Eight studies included both episodic and chronic migraine patients (16,17,25–27,29,30,39), comprising a total of 527 episodic and 258 chronic migraine patients (Supplementary Table 2). The pooled relative frequency of neck pain was 77.0% (95% CI: 69.0–84.0; I2 = 72.0%) in episodic migraine patients and 87.0% (95% CI: 77.0–93.0; I2 = 61.4%) in chronic migraine patients (Supplementary Table 2) (Figure 4). The odds of neck pain were two times higher in chronic migraine patients compared to episodic migraine patients (OR: 2.04, 95% CI: 1.3–3.3; P = 0.008; I2 = 10.0%) (Figure 3).

Likelihood of Neck Pain in Patients with Migraine Compared to Healthy Controls.

Likelihood of Neck Pain in Patients with Chronic Migraine vs. Episodic Migraine.

Mean Relative Frequency of Associated Neck Pain in Patients with Migraine, Non-Headache Controls, Episodic and Chronic Migraine Patients in Clinic-based Sample. (A) The diagram illustrates the pooled relative frequencies of neck pain in migraine patients (blue bar) and non-headache controls (green bar). Each bar includes a 95% confidence interval (black line). (B) The diagram illustrates the pooled relative frequencies of neck pain in patients with episodic (orange bar) and chronic migraine (red bar). Each bar includes a 95% confidence interval (black line).
Timing of neck pain
Six studies described neck pain in relation to the migraine phases (2,18,23,31–33). Neck pain was reported as a migraine trigger among 31.9% and 57.0% migraine patients in two studies (23,31). In one study (n = 487), 37.8% of patients experienced neck pain upon onset of the migraine attack, 24.2% experienced neck pain within 2 hours before the onset of attack (neck pain continued into the attack state in 82.2% of these cases), while 7.4% experienced neck pain 2–48 hours before onset of attack (neck pain continue into the attack state in 22.2% of these cases) (2). Another study (n = 100) reported higher neck pain incidence in migraine patients with 85.0% reporting neck pain before (2–48 hours before), 90.0% during and 37.0% after attacks (18). Three studies reported that migraine attacks and neck pain started and ended concurrently with varying incidences at 64.0% [21], 68.1% (31), and 89.1% (33).
Population-based studies
Two population-based studies (3,40) were identified, and thus, analyses were not performed. One study reported one-year coexistent neck pain prevalence of 76.2% for pure migraine patients (n = 80), while the prevalence was 88.4% for pure TTH patients (n = 138), compared to 56.7% for healthy participants (n = 476) (3). The neck pain prevalence was slightly higher at 89.3% in patients with coexistent migraine and TTH (103). Coexistent neck pain was more prevalent among TTH compared with migraine patients, but when adjusting for age, adjusted for age, gender, education and poor self-rated health, the difference was no longer significant (p = 0.066) (3). The study also performed a logistic model, which examined headache types as a predictor of neck pain. After adjusting for age, gender, education and poor self-related health, the relative odds of neck pain in patients with primary headaches (i.e., pure TTH or coexistent migraine and TTH) were four times higher than in non-headache controls (p = 0.001).
The other population-based study investigated headache among children (15–18 years of age) from 11 public grammar schools in Munich (40). The study included 378 students with migraine, 614 with TTH and 213 students without headache. Patients were asked to mark specific regions of muscular pain in schematic drawings. Associated neck pain was a prominent symptom, reported by 38.9% of students with migraine (p < 0.001), 22.6% of students with TTH (p < 0.01) compared to 14.1% of students without headache.
Study assessment
The quality of each study included in this systematic review and meta-analysis was critically evaluated using a critical appraisal checklist for prevalence studies from the Joanna Briggs Institute (JBI). One study was deemed low quality as it only scored 3 “yes” out of 9 on the quality assessment scale (32). Two other studies were demeaned moderate quality as they scored 6 “yes” out of 9 quality assessment scale (2,37). The remaining 21 studies had a very low risk of bias (Supplementary Table 3).
Discussion
The meta-analyses demonstrated that neck pain is very frequent in migraine patients compared to non-headache controls (77.0% vs. 23.2%) in clinic-based studies. A comparable neck pain prevalence at 76.2% was reported in the only population-based study of adult migraine patients (3). One of the most striking findings was that neck pain was ∼12 times more likely in migraine patients compared to non-headache controls in clinic-based studies. Our study confirms that neck pain is a prominent feature of migraine (12,13) and potentially an important component in migraine pathophysiology.
Neck pain was found to be two-times more likely in patients with chronic migraine compared to episodic migraine. This finding is supported by a positive correlation between migraine attack frequency and neck pain from both population- and clinic-based studies (3,14,16,17,22,41). Increased attack frequency could potentially lead to an increased responsiveness of wide dynamic range neurons in the trigeminocervical complex, which may lower attack threshold, and lead to central sensitization (10). In this case, neck pain is merely a secondary phenomenon of migraine attacks. However, we recently found no difference in the mechanical pain threshold between chronic migraine patients with and without on-going headache (42). The magnitude of the mechanical pain threshold was comparable to reports in non-headache controls (similar sex and age group) using similar methodology (43). In addition, one previous study reported no difference in the mechanical and thermal pain thresholds between episodic and chronic migraine patients (44). Altogether, these findings question whether central sensitization alone explains the increased frequency of neck pain in chronic migraine patients compared to episodic migraine patients.
An alternative explanation could be that local inflammation in neck muscles leads to sensitization of neck muscle nociceptor (45). The question is whether this results in sensitization of trigeminal nociceptors, surrounding the meninges vasculature, and promotion of attacks. Indeed, a functional connection between intracranial and extracranial structures has been reported (46,47) as peripheral nociceptive fibers (from C2 dorsal root ganglion) traverse the occipital muscles before innervating the posterior dura (46). Furthermore, biopsies from the neck periosteum in chronic migraine patients with neck pain demonstrated increased expression of pro-inflammatory genes and decreased expression of anti-inflammatory genes (48).
In relation to the migraine phases, five out of six studies suggest that onset of neck pain is frequent within the headache phase of migraine attacks (range 37.8%–90.0%) (2,18,31–33). Two studies also reported onset of neck pain before the headache phase of migraine (range 7.4%–85.0%) (2,18), and neck pain was reported as a trigger of the migraine as well (range 31.9%–57.0%) (23,31). Thus, further experimental studies are needed to clarify whether neck pain is a consequence or cause of the migraine attack.
Limitations
Our meta-analyses demonstrated some heterogeneity (I2 from 61.3% to 72.0%) among studies. Consequently, our pooled neck pain relative frequency estimates should be interpreted with caution. The heterogeneity may arise from several factors. First, the clinic-based studies used either prospective (n = 22) or retrospective (n = 2) approaches. The longitudinal design of prospective studies carries the risk of loss to follow up, while retrospective studies may be influenced by recall bias. Second, 13 out of the 24 clinic-based studies included both sexes (2,15,18,22,23,32–38,49). Here, the proportion of female participants was expectedly higher than male participants. The remaining 10 studies included only female participants (see Table 2). Thus, the overall sex distribution in the 24 studies do not represent the general sex distribution in migraine. Hence, it is unclear whether our results would be replicated in larger representative sample that would include more men with migraine. Third, the definition of neck pain was only stated in three clinic-based studies (2,31,39), while unspecified in the remaining studies. Fourth, neck pain was self-reported in 10 clinic-based studies (15,22,25–30,32,37) and both population-based studies (3,40) without the use of cut-off scores or clear-cut definitions. However, 14 clinic-based studies used either Neck Disability index (14,16,17,24,34–36,38,39) or interviews (2,18,23,31,33) to assess neck pain. Fifth, only two studies (clinic-based) used daily headache diary to assess the migraine diagnosis (15,18). This may influence the correct migraine diagnosis (e.g., episodic vs. chronic migraine). Sixth, only one of the clinic-based studies reported if the included migraine patients also experienced TTH (39). This lack of information is a major limitation, as pericranial tenderness in TTH may contribute to higher relative frequency of neck pain than in migraine (50). This is supported by the one population-based study in adults, which reported relatively more frequent neck pain among patients with both migraine and TTH than those with pure migraine (89.0% vs. 76.0%) (3). Seventh, only six clinic-based studies reported on the timing of neck pain in relation to migraine, with varied outcomes (2,18,23,31–33). At present, we cannot state the exact time to onset of the neck pain in relation to the migraine phases, although the few available studies suggest neck pain to be most common during the headache phase (2,18,31–33). Finally, all except two clinic-based studies (2,32) included migraine patients from outpatient clinics, where this subgroup of migraine patients might be more severely affected than others. This limitation represents an overestimation of the neck pain prevalence in the migraine population. It is likely that the reported migraine and neck pain severity is higher compared with real world data, as majority of the qualified studies had a clinic-based approach. Indeed, neck pain prevalence is relatively higher within comorbid conditions, such as fibromyalgia, whiplash-associated disorder, rheumatic disease and degenerative spine disease (51).
Future directions
Future research should take into consideration the methodological limitations presented here. More studies are needed to provide a better estimate of the neck pain prevalence in the migraine population. We suggest that future studies should clearly define neck pain and apply validated instruments, such as the Neck Disability index. Moreover, it is important that up-coming studies report the onset of neck pain in relation to the migraine phase and all data is collected in a prospective fashion using diaries. In this context, it is interesting to shed light on if neck pain before the onset of migraine attack differs from the neck pain that appears after attack onset. It would also be interesting to investigate possible differences in neck prevalence between sexes and between those with and without migraine aura. Finally, prevalence of co-existing TTH in migraine patients should also be clearly reported in future studies, allowing subgroup analysis between those with both migraine and TTH vs pure migraine. In up-coming comparison studies between migraine and TTH patients, regarding neck pain, it is crucial to illuminate the following: 1) is neck pain a cause or consequence of migraine and TTH; 2) is neck pain more frequent in TTH or migraine patients; 3) are there migraine or TTH symptoms which only appear in patients with neck pain, but not in patients without neck pain.
Conclusion
The present meta-analyses show that neck pain is ∼12 times more likely in migraine patients compared to non-headache controls in clinic-based studies. Moreover, neck pain is two times more frequent in patients with chronic migraine compared to episodic migraine in clinic-based studies. Neck pain seems to be present in all stages of the migraine attack, but the majority of studies indicate that neck pain is most common during the headache phase of a migraine attack. While our meta-analyses demonstrate considerable heterogeneity between the included studies, it also highlights important aspects to consider in future research of neck pain in migraine to improve our understanding of the driving mechanisms of neck pain in a major group of migraine patients.
Key finding
The meta-analysis from 8 clinic-based studies demonstrated a pooled relative frequency of neck pain of 77.0% (95% CI: 69.0–86.4) in patients with migraine and 23.2% (95% CI:18.6–28.5) in non-headache controls. Neck pain is more frequent in patients with chronic migraine (87.0%, 95% CI: 77.0–93.0) compared to episodic migraine (77.0%, 95% CI: 69.0–84.0). Neck pain is 12 times more prevalent in patients with migraine patients compared to non-headache controls and two times more prevalent in patients with chronic migraine compared to episodic migraine.
Supplemental Material
sj-pdf-1-cep-10.1177_03331024211068073 - Supplemental material for Prevalence of neck pain in migraine: A systematic review and meta-analysis
Supplemental material, sj-pdf-1-cep-10.1177_03331024211068073 for Prevalence of neck pain in migraine: A systematic review and meta-analysis by Haidar Muhsen Al-Khazali, Samaira Younis, Zainab Al-Sayegh, Sait Ashina, Messoud Ashina and Henrik W Schytz in Cephalalgia
Footnotes
Author Contributions
HM Al-Khazali, S Younis and HW Schytz conceived and designed (including search strategies) the review. HM Al-Khazali did the literature search with Z Al-Sayegh. HM Al-Khazali wrote the first and subsequent drafts of the manuscript. S Younis, HW Schytz, Z Al-Sayegh, M Ashina and S Ashina participated in critical revision and writing of the article. All authors have seen and approved the final version.
Conflict of interest
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: HWS received speaking fees from Novartis. SA has received honoraria from Allergan and Teva and consultant fees from Amgen and Allergan. SA received honoraria for lecturing from Allergan and Avanir Pharmaceuticals. MA has received personal fees from Alder BioPharmaceuticals, Allergan, Amgen, Eli Lilly, Novartis, and Teva. MA also participated in clinical trials as the principal investigator for Alder, Amgen, electroCore, Novartis, and Teva. MA also serves as an associate editor Federation. MA reports research grants from Lundbeck Foundation, Research Foundation of the President-elect of the International Headache Society, and General Secretary of the European Headache Capital Region of Copenhagen, and Novo Nordisk Foundation. The other authors declare no conflicts of interest.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
