Investigation of the interaction between shoulder and cervical vertebral pathologies in isolated shoulder pain: clinical and morphological aspects

Abstract

Background

The anatomical interaction and overlap between the cervical vertebra and the shoulder presents a diagnostic challenge for both orthopedic surgeons and neurosurgeons. Musculoskeletal disorders, affecting 1.71 billion people globally, often involve shoulder pain, which is frequently linked to cervical issues. Understanding this connection is crucial for accurate diagnosis and treatment. This study aimed to explore potential structural associations between annular bulging and shoulder pathologies in patients presenting with isolated shoulder pain.

Methods

Using a retrospective screening approach, we analyzed MRI images of patients admitted to the Orthopedics Clinic between 2017 and 2019 who presented with pain in the shoulder region, identified through the Hospital Management Information System. A total of 618 patients were included, and their demographic details, along with shoulder and cervical pathologies, were recorded and assessed. The association between annular bulging and shoulder pathologies was evaluated using regression models as sex and age covariants.

Results

In this study of 618 patients with isolated shoulder pain, 61.3% were female (n = 379), and the mean age of the patient group was 51.66 ± 9.5 years. Pathological findings appeared in 47.7% of cases in both cervical spine and shoulder MRIs, with 24.1% showing pathology only in shoulder MRIs and 5.7% only in cervical MRIs. Annular bulging was most common, especially in the C5-C6 segment. Women exhibited a higher incidence of acromion type 2 morphology and subcoracoid effusion, while men commonly had subcoracoid effusion and acromioclavicular joint arthrosis.

Conclusions

This exploratory study demonstrates morphological associations between cervical spine and shoulder pathologies in patients presenting with shoulder pain. It is imperative to interpret these findings as descriptive structural correlations rather than causal relationships. The observed co-occurrence is likely to reflect shared degenerative and biomechanical processes affecting adjacent anatomical regions.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-026-07011-y.

Introduction

Global Burden of Disease 2019 data reveals that around 1.71 billion people globally suffer from musculoskeletal disorders, including fractures and injuries [1]. The prevalence of these conditions varies by age and specific diagnoses, affecting individuals globally, with shoulder pain showing considerable variability across countries, with a median occurrence of 16% (ranging from 0.67 to 55.2%) [2].

The etiology of shoulder pain encompasses a broad spectrum, incorporating local pathologies within the shoulder, including the glenohumeral joint, acromioclavicular joint, sternoclavicular joint, rotator cuff, and various soft tissues within the shoulder complex [3, 4]. Additionally, it may emanate as referred pain from extrinsic sources, such as the cervical region or fascia [3, 5]. Although surgery rates are rising in some regions, numerous studies on shoulder impingement syndrome indicate no difference in pain and disability outcomes between surgical and conservative treatments. However, outcomes vary considerably across different shoulder conditions and individual patient factors. Studies have shown that approximately 50% of new shoulder pain cases fully recover after six months, and 60% after twelve months [6, 7].

In the literature, several studies on the cervical column are concerned with general explanations or specific relationships between head posture, craniofacial morphology, and upper cervical vertebral osseous anatomy. Although there are publications explaining stability disorders in bony structures of the shoulder region, the muscles attached to them, and explaining radiculopathy that associates the cervical spinal nerves innervating the muscles, there are limited studies correlating the cervical vertebral column anatomical changes with shoulder pain.

Research indicates that approximately 40% of presentations involving extremities had symptoms identified as originating from the spine and exhibited positive responsiveness to spinal interventions [8]. The cervical spine and shoulder exhibit shared somatic referral patterns, leading to potential overlap in symptoms, resulting in mixed manifestations of shoulder pain [9].

Understanding whether there exists an association between isolated shoulder pain and cervical pathologies is crucial for accurate diagnosis, effective treatment, and improved patient outcomes.

The primary aim of this study was to investigate the relationship between annular bulging in the cervical regions and isolated shoulder pain by evaluating anatomical structures and pathologies in patients presenting with shoulder pain. Additional cervical findings identified by imaging were recorded and reported descriptively to characterize the broader pathological profile of the study group; however, in accordance with the predetermined primary analysis, formal hypothesis testing was limited to annular bulging.

Material and methods

This study was approved by the Clinical Research Ethics Committee of the affiliated university under the ethical standards laid down in the 1964 Declaration of Helsinki (2020-06-24-461).

A total of 618 patients who presented to the orthopedic clinic and were seen by the first author between the years 2017 and 2019 were included in our study. The cervical region of the patients was also evaluated by a neurosurgeon.

The concurrent MRI assessment encompassing both cervical and shoulder regions within a singular session, led by the first author, was undertaken with the objective of refining the differential diagnosis of cervical pathologies in the context of patient admissions. We retrospectively examined the magnetic resonance imaging (MRI) images of patients admitted in the Hospital Management Information System who presented with shoulder pain, regardless of whether their diagnosis was subsequently confirmed.

Inclusion criteria were

(1) Patients presenting to the clinic with complaints of shoulder pain persisting for at least two months; (2) Age 18 years and older; (3) Concomitant acquisition of cervical and shoulder MRI scans within a single session.

Exclusion criteria were

(1) Patients with neck pain; (2) Patients with pain radiating to the entire arm, forearm, and wrist; (3) Patients with back pain; (4) Patients with pain in the contralateral shoulder; (5) Patients with accompanying upper extremity nerve compression (carpal tunnel syndrome, ulnar tunnel neuropathy, etc.); (6) Patients who have undergone shoulder surgery or trauma; (7) Patients diagnosed with rheumatic diseases or any musculoskeletal system diseases affecting the upper extremity, humeral head osteonecrosis, chronic steroid use, or sickle cell disease.

Patients meeting the inclusion and exclusion criteria were included in the study for evaluation (Fig. 1).

Fig. 1.

PRISMA-style flow diagram illustrating the selection of study population

Cervical and shoulder MRI images were examined retrospectively using the PACS (Picture Archiving and Communication System) by an experienced radiologist, an orthopedist, and a neurosurgeon. The MRI images of both the shoulder and cervical regions from the patients were stratified into two categories: pathology present and pathology absent. Within the context of MRI reporting, the presence of effusion and signal alterations was acknowledged as indicative of pathology. The shoulder and cervical pathologies were evaluated, and it was investigated whether there was any association between them. The primary cervical variable examined was the annular bulging, analyzed relative to the cervical level. All other cervical pathologies detected on MRI were systematically recorded but designated as secondary descriptive variables prior to analysis. Because their frequencies were insufficient for a stable formal inference and their findings fell outside the scope of the predetermined primary hypothesis, these findings are reported only as descriptive statistics, and no formal hypothesis testing was performed for these variables. All measurements were performed blinded to subject details and data of any other results.

It should be noted that acromion morphology (Types 1, 2, and 3) represents anatomical variants rather than pathological conditions. However, given the proposed association of acromion type with rotator cuff pathology in some literature, we included this morphological classification in our analysis as a potentially relevant anatomical factor.

Additionally, neural foraminal narrowing was not evaluated as a separate variable; instead, it was assessed within the osteophyte category, as the foraminal narrowing observed in this study was predominantly secondary to osteophytic changes.

Statistical analysis

The IBM SPSS 26.0 (IBM Corp., Armonk, NY) package program and MS Excel 2019 were used to evaluate the data. Descriptive statistical variables such as age and gender were expressed as frequency/number (n). A single primary hypothesis was pre-specified prior to analysis: the relationship between annular bulging at different cervical levels and co-existing shoulder pathologies. In order to account for potential confounding, multivariable logistic regression was performed for all primary comparisons. In each model, age and sex were entered as covariates. The Benjamini–Hochberg procedure was applied to control the false discovery rate (FDR) at 5%. Both unadjusted p values and FDR-adjusted q-values are reported for all comparisons within the primary analysis. Sparse cell counts were observed in some pathological categories, which hindered the consistency of statistical inferences. These categories were then transformed into clinically significant composite endpoints based on pre-defined clinical rationale, or – where combination was not feasible – presented as purely descriptive statistics.

To measure the strength and direction of the associations between cervical and shoulder pathologies, Odds Ratios (OR) with 95% Confidence Intervals (CI) were calculated. Furthermore, a post-hoc power analysis was performed using G*Power (version 3.1.9.6) to evaluate the adequacy of the sample size in detecting significant clinical associations. A p value less than 0.05 was considered statistically significant.

A sensitivity analysis was conducted restricting the sample to participants aged ≥ 60 years to evaluate whether associations were disproportionately driven by age-related degenerative burden. For effect size interpretation, findings were prioritized according to both statistical precision and magnitude; associations with adjusted OR > 2.0 and narrow 95% confidence intervals were considered the most robust and are discussed preferentially. Associations with OR < 2.0 or wide confidence intervals are reported for completeness but interpreted with appropriate caution.

Results

In our study, 618 individuals who were admitted to the orthopedic clinic with the complaint of isolated shoulder pain and met the inclusion and exclusion criteria were screened retrospectively. The primary analysis examined the association between annular bulging at each cervical level and the presence of shoulder pathologies. Additional cervical pathologies identified during imaging review are summarized descriptively in Figs. 2 and 3. These findings were not subject to formal hypothesis testing and are reported to characterize the full imaging spectrum of the study cohort.

Fig. 2.

Heatmaps illustrating the distribution of the most common cervical pathologies, stratified by sex and age group. Panels A–B represent Female and Male participants, respectively, while panels C–F show the corresponding age-group distributions across both sexes: < 40 years (C), 40–49 years (D), 50–59 years (E), and ≥ 60 years (F). Within each panel, rows represent anatomical regions and columns represent individual pathologies. Color intensity indicates the number of individuals (n) presenting with each pathology. An identical color scale was applied across all panels to enable direct comparisons

Fig. 3.

Clustered bar charts showing the most common shoulder pathologies, stratified by sex and age group. Panel A depicts the distribution by sex (Female vs. Male), while Panel B shows the distribution across age groups: < 40 years, 40–49 years, 50–59 years, and ≥ 60 years. Each bar represents the number of individuals (n) with the corresponding pathology

Patient demographics

Of the 618 patients, 239 were male and 379 were female, with a mean age of 51.66 ± 9.5 years. The age distribution was as follows: 107 (17.3%) patients were under 40 years, 146 (23.6%) were between 40 and 49 years, 186 (30.1%) were between 50 and 59 years, and 179 (29.0%) were 60 years and above.

No pathological findings were found in the shoulder and cervical MRI images of 21.2% (n = 131) of patients. In 47.7% (n = 295) of patients, pathology was detected in both cervical spine and shoulder MRI images. In 24.1% (n = 149) of patients, pathology was seen only on shoulder MRI images. In 5.7% (n = 35) of patients, pathology was seen only on cervical MRI images.

Most common cervical pathologies

The cervical pathologies among the patients were categorized. The most common pathology for both men and women in all segments was annular bulging (Fig. 4). It should be noted that annular bulging is highly prevalent in the general population and may not be directly symptomatic; however, we report these findings as observed on MRI.

Fig. 4.

Cervical pathologies. A C3-C4 annular bulging. B C4-C5 annular bulging. C C5-C6 annular bulging (black arrow), C6-C7 annular bulging (white arrow). Note: Interpretation of these MRI findings should be made in clinical context, as annular bulging is highly prevalent in the asymptomatic population

The highest number of pathologies in the C2-C3 segment was seen in the 50–59 age range (n = 3), and the most common cervical pathology was annular bulging in both women (n = 4) and men (n = 2). The highest number of pathologies was seen in the 60 years and above age group for C3-C4 (n = 41), C4-C5 (n = 48), C5-C6 (n = 77), C6-C7 (n = 63), and C7-T1 (n = 6) segments.

The most common cervical pathology among all cervical segments was annular bulging in both women and men, with the C5-C6 segment showing the highest prevalence (women n = 122, men n = 78).

Most common shoulder pathologies

The most common pathology in the shoulder was rotator cuff pathologies (Figs. 5 and 6). According to tear types, the most common tear type in both women (n = 87) and men (n = 37) was the partial tear (bursal side) of supraspinatus. It was observed mostly in the 60 years and above (n = 55) age group. Partial tear of the infraspinatus was the second most common tear type in both women (n = 39) and men (n = 13), observed mostly in the 60 years and above (n = 27) age group. In the subscapularis, partial tear was most common in both women (n = 12) and men (n = 4), observed mostly in the 60 years and above (n = 11) age group.

Fig. 5.

Rotator cuff pathologies. A Supraspinatus muscle total rupture. B Infraspinatus muscle rupture. C Subscapularis muscle rupture

Fig. 6.

Supraspinatus muscle pathologies. A Supraspinatus muscle bursal side partial tear. B Supraspinatus muscle PASTA tear. C Supraspinatus tendinosis

Subcoracoid effusion was the most common shoulder finding other than rotator cuff pathologies in both sexes (n = 182 for women, n = 86 for men). It was seen mostly in the age group of 60 years and above (n = 102). Subcoracoid effusion was followed by acromion type 2 morphology in women (n = 228) and acromioclavicular joint arthrosis (n = 72) in men.

Association between cervical and shoulder pathologies

Multivariate logistic regression was performed for all primary comparisons, adjusting for age and sex. FDR correction was applied to all data. Relationships with adjusted odds ratio (aOR) > 2.0 and narrow confidence intervals were prioritized as the most robust estimates. Full results for all comparisons, including non-significant relationships, are presented in Table 1, and the clinically most significant findings for aOR and CI are visualized in a Forest Plot (Fig. 7). To increase transparency, an additional table presenting the fully adjusted logistic regression models for all primary comparisons is provided (Supplementary Table 1).

Table 1.

Multivariable logistic regression analysis of the association between cervical annular bulging and shoulder pathologies, adjusted for age and sex

	Annular bulging
	C3-C4		C4-C5		C5-C6		C6-C7
	N	Raw p (pFDR) aOR(CI)	N	Raw p (pFDR) aOR(CI)	N	Raw p (pFDR) aOR(CI)	N	Raw p (pFDR) aOR(CI)
Supraspinatus muscle pathologies	44	0.709 (0.879) 1.11 (0.64 – 1.95)	65	0.879 (0.879) 0.96 (0.57 – 1.61)	111	0.000 (0.000) 2.50 (1.58–3.95)	87	0.152 (0.405) 1.41 (0.88–2.25)
Infraspinatus muscle pathologies	29	0.830 (0.879) 1.07 (0.57 – 2.01)	34	0.835 (0.879) 1.07 (0.59 – 1.93)	55	0.080 (0.320) 1.63 (0.94–2.83)	45	0.227 (0.454) 1.40 (0.81–2.40)
ACJ arthrosis	41	0.261 (0.298) 1.33 (0.81–2.18)	64	0.012 (0.019) 1.73 (1.13–2.66)	97	0.002 (0.004) 1.82 (1.25–2.64)	76	0.029 (0.041) 1.56 (1.05–2.32)
ACJ effusion	24	0.003 (0.006) 2.31 (1.33–4.00)	26	0.209 (0.271) 1.40 (0.83–2.34)	37	0.342 (0.342) 1.25 (0.79–2.00)	31	0.274 (0.299) 1.31 (0.81–2.14)
Acromion Type 2	68	0.002 (0.005) 2.46 (1.41–4.28)	99	0.000 (0.000) 2.80 (1.73–4.54)	164	0.000 (0.000) 4.29 (2.81–6.54)	123	0.000 (0.000) 2.59 (1.69–3.98)
Subacromial bursitis	38	0.299 (0.312) 1.29 (0.80–2.11)	54	0.253 (0.304) 1.28 (0.84–1.96)	95	0.000 (0.000) 1.99 (1.38–2.88)	82	0.000 (0.000) 2.34 (1.58–3.46)
Subdeltoid bursitis	24	0.019 (0.029) 1.94 (1.11–3.38)	28	0.203 (0.264) 1.40 (0.83–2.35)	47	0.008 (0.014) 1.87 (1.18–2.96)	40	0.005 (0.009) 1.97 (1.22–3.16)
Subcoracoid effusion	57	0.001 (0.003) 2.36 (1.43–3.90)	83	0.000 (0.000) 2.61 (1.70–4.01)	122	0.000 (0.000) 2.38 (1.65–3.43)	110	0.000 (0.000) 3.63 (2.42–5.46)

aOR, adjusted odds ratio; CI, confidence interval. All models were adjusted for age and sex. FDR correction was applied across all comparisons within the primary analysis

Bold values indicate statistically significant findings prioritized based on effect size (approximately OR > 2.0) with relatively narrow 95% confidence intervals

Fig. 7.

Forest plot showing the Odds Ratios and 95% Confidence Intervals for the significant associations between cervical annular bulging and shoulder pathologies. The vertical dashed line indicates the null effect (OR = 1)

Subsequent to FDR correction, annular bulging of the C3-C4 region was found to be significantly associated with three distinct shoulder findings. The strongest association was observed with ACJ effusion (aOR 2.31, 95% CI 1.33–4.00; p = 0.003, q = 0.006), though the relatively wider confidence interval for this estimate reflects the lower absolute frequency of ACJ effusion in the sample and limits point estimate precision. The findings of this study indicate that type 2 acromion morphology (aOR 2.46, 95% CI 1.41–4.28; p = 0.002, q = 0.005) and subcoracoid effusion (aOR 2.36, 95% CI 1.43–3.90; p = 0.001, q = 0.003) are significantly associated with the outcome, with effect sizes exceeding 2.0 in both cases. Subdeltoid bursitis demonstrated a nominally significant association (aOR 1.94, 95% CI 1.11–3.38; p = 0.019, q = 0.029) with a point estimate approaching but not exceeding 2.0. No significant associations were identified between C3-C4 annular bulging and supraspinatus or infraspinatus muscle pathologies, ACJ arthrosis, or subacromial bursitis at this level.

C4-C5 annular bulging was linked to three shoulder pathologies. Type 2 acromion morphology had the largest and most precise effect at this level (aOR 2.80, 95% CI 1.73–4.54; p < 0.001, q < 0.001). Subcoracoid effusion was also linked to a significant effect (aOR 2.61, 95% CI 1.70–4.01; p < 0.001, q < 0.001). ACJ arthrosis was also significant after FDR correction, but with a smaller effect (aOR 1.73, 95% CI 1.13–2.66; p = 0.012, q = 0.019). Supraspinatus and infraspinatus muscle pathologies, ACJ effusion, subacromial bursitis, and subdeltoid bursitis did not reach significance at this level following FDR correction.

C5-C6 annular bulging demonstrated the broadest and strongest association profile across all cervical levels examined, reaching significance for six shoulder pathologies after FDR correction. The largest effect was observed with Type 2 acromion morphology (aOR 4.29, 95% CI 2.81–6.54; p < 0.001, q < 0.001), representing the single strongest association in the entire dataset. Subcoracoid effusion (aOR 2.38, 95% CI 1.65–3.43; p < 0.001, q < 0.001) and supraspinatus muscle pathologies (aOR 2.50, 95% CI 1.58–3.95; p < 0.001, q < 0.001) also demonstrated effect sizes exceeding 2.0 with narrow confidence intervals. ACJ arthrosis (aOR 1.82, 95% CI 1.25–2.64; p = 0.002, q = 0.004), subacromial bursitis (aOR 1.99, 95% CI 1.38–2.88; p < 0.001, q < 0.001), and subdeltoid bursitis (aOR 1.87, 95% CI 1.18–2.96; p = 0.008, q = 0.014) were additionally significant, with effect sizes approaching but not exceeding 2.0. Infraspinatus muscle pathologies did not reach significance at C5-C6 after FDR correction (aOR 1.63, 95% CI 0.94–2.83; q = 0.320).

C6-C7 annular bulging was linked to five shoulder pathologies after FDR correction. The strongest link was with subcoracoid effusion (aOR 3.63, 95% CI 2.42–5.46; p < 0.001, q < 0.001), the second largest point estimate in the entire dataset. Type 2 acromion morphology (aOR 2.59, 95% CI 1.69–3.98; p < 0.001, q < 0.001) and subacromial bursitis (aOR 2.34, 95% CI 1.58–3.46; p < 0.001, q < 0.001) had effect sizes over 2.0 with narrow confidence intervals. Subdeltoid bursitis (aOR 1.97, 95% CI 1.22–3.16; p = 0.005, q = 0.009) and ACJ arthrosis (aOR 1.56, 95% CI 1.05–2.32; p = 0.029, q = 0.041) were also significant, with the latter having the smallest effect size among all FDR-corrected significant findings and warranting cautious interpretation. Supraspinatus muscle pathologies, infraspinatus muscle pathologies, and ACJ effusion did not reach significance at this level.

We performed a limited sensitivity analysis with participants ≥ 60 years of age (Table 2).

Table 2.

Sensitivity analysis restricted to participants aged ≥ 60 years: multivariable logistic regression of cervical annular bulging and shoulder pathology associations

	Annular bulging
	C3-C4		C4-C5		C5-C6		C6-C7
	N	Raw p (pFDR) aOR(CI)	N	Raw p (pFDR) aOR(CI)	N	Raw p (pFDR) aOR(CI)	N	Raw p (pFDR) aOR(CI)
Supraspinatus Muscle Pathologies	28	0.050 (0.050) 2.17 (1.00–4.70)	38	0.002 (0.003) 3.54 (1.61–7.77)	59	0.000 (0.000) 4.17 (2.14–8.14)	48	0.003 (0.004) 2.87 (1.44–5.72
Infraspinatus Muscle Pathologies	16	0.184 (0.283) 1.69 (0.78–3.69)	20	0.283 (0.283) 1.52 (0.71–3.28)	30	0.170 (0.283) 1.60 (0.82–3.13)	26	0.245 (0.283) 1.51 (0.75–3.01)
ACJ arthrosis	32	0.001 3.85 (1.68–8.78)	36	0.005 2.91 (1.37–6.16)	54	0.001 2.87 (1.52–5.40)	43	0.023 2.15 (1.11–4.18)
ACJ effusion	15	0.006 3.08 (1.38–6.89)	16	0.039 2.34 (1.04–5.25)	18	0.332 1.45 (0.68–3.09)	14	0.622 1.22 (0.55–2.69)
Acromion Type 2	38	0.001 8.50 (2.48–29.22)	43	0.001 5.37 (2.03–14.22)	67	0.000 5.99 (2.74–13.10)	50	0.020 2.40 (1.15–5.04)
Subacromial bursitis	23	0.124 1.77 (0.86–3.66)	27	0.128 ( 1.73 (0.85–3.51)	45	0.007 2.35 (1.26–4.37)	39	0.005 2.55 (1.32–4.92)
Subdeltoid bursitis	15	0.008 2.95 (1.33–6.59)	14	0.199 1.71 (0.75–3.90)	21	0.089 1.90 (0.91–4.01)	15	0.708 1.16 (0.54–2.49)
Subcoracoid effusion	32	0.003 3.54 (1.55–8.11)	40	0.000 4.99 (2.18–11.46)	55	0.002 2.81 (1.49–5.32)	49	0.000 3.95 (1.93–8.10)

aOR, adjusted odds ratio; CI, confidence interval. All models were adjusted for age and sex. FDR correction was applied across all comparisons within the primary analysis

Bold values indicate statistically significant findings prioritized based on effect size (OR > 2.0) with relatively narrow 95% confidence intervals

In the ≥ 60 s sensitivity analysis, associations changed in both size and location. Effect sizes were higher, and several relationships became more consistent across cervical segments. Supraspinatus pathology demonstrated stronger, more widespread associations in this subgroup. ACJ arthrosis showed increased effect sizes compared to the main model, supporting a greater contribution of degenerative joint changes with age. Subacromial bursitis also exhibited a more pronounced pattern, extending beyond the lower cervical levels observed in the primary analysis. Acromion type II and subcoracoid effusion remained the most robust findings in both analyses, maintaining strong and consistent associations across all levels.

Overall, the sensitivity analysis suggests that age-related degenerative burden increases the strength of associations and expands their segmental distribution, particularly for rotator cuff–related and degenerative shoulder pathologies.

Discussion

The morphological interaction and overlap between the cervical vertebra and the shoulder poses a challenge for both orthopedic surgeons and neurosurgeons because the symptoms of vertebral and shoulder pathologies are often similar and may lead to misdiagnosis. Previous research has suggested that cervical vertebral surgery may affect shoulder function and reveal shoulder pathologies that were not apparent before surgery, making both preoperative and postoperative shoulder examinations essential. Additionally, cervical spine pathology and surgery can negatively impact the outcome of subsequent shoulder surgery such as rotator cuff repair or shoulder arthroplasty [10].

Kholinne et al. declared that cervical disorders causing neck pain may mask serious shoulder pathologies like rotator cuff disease and subsequent adhesive capsulitis as well as calcific tendinitis. However, these coexisting issues can be identified with a careful preoperative examination [11].

A study by Sembrano et al. estimated that approximately 3.6–4% of cases with shoulder or neck symptoms may exhibit alternation between shoulder and neck symptoms [12]. This is consistent with our finding that 5.7% of patients with isolated shoulder pain had cervical pathologies without concurrent shoulder pathology.

In the present study, our findings demonstrate statistical associations between multiple shoulder pathologies in patients with isolated shoulder pain. Pathology was detected in both cervical spine and shoulder MRI images in 47.7% of patients, while in 24.1% of patients, pathology was seen only on shoulder MRI images.

Interpretation of associations

It is important to interpret these associations with caution. The high prevalence of concurrent cervical and shoulder pathology (47.7%) may partly reflect the fact that both cervical degenerative changes (particularly annular bulging) and shoulder pathologies are extremely common in the general population, including asymptomatic individuals. Nakashima et al. reported that annular bulges are found in approximately 88% of the asymptomatic population in cervical MRI studies [13]. Similarly, rotator cuff abnormalities are frequently identified in asymptomatic shoulders on MRI.

Therefore, our findings of statistical associations should not be interpreted as evidence of causation. The co-occurrence of cervical and shoulder pathology in our symptomatic population may reflect shared risk factors such as age-related degeneration, occupational factors, or genetic predisposition, rather than direct pathophysiological relationships.

Rotator cuff injuries are the most prevalent upper extremity condition encountered by primary care physicians and orthopedic surgeons. These injuries are influenced by non-modifiable risk factors such as age, genetics, and family history, and modifiable factors like diabetes, smoking, body mass index, and occupational hazards [14]. Importantly, the supraspinatus tendon has a well-documented hypovascular critical zone that predisposes it to degeneration and tears independent of cervical spine factors. Our study found rotator cuff pathologies to be the most common shoulder conditions, identifying 16 different pathologies.

The data revealed a higher prevalence of shoulder pain in middle-aged and older individuals, aligning with literature on age-related musculoskeletal issues, and showed that females (61.3%) reported shoulder pain more frequently than males [15].

In our study, sensitivity analyses performed on patients over 60 years of age showed higher effect sizes and more consistent relationships across cervical segments. We found that supraspinatus pathology showed stronger and more widespread associations in this subgroup.

There is evidence suggesting that pain occurring in the upper extremities, such as the shoulder and arm, may originate from pathologies in the cervical region [16, 17]. Noh and Park [18] reported that degenerative changes in the cervical spine can lead to nerve root compression and pain radiating to the upper extremities. One of the most evident examples of this is radiculopathy, characterized by pain extending from the neck to the shoulder, arm, and even the hand, leading to sensory and motor impairment [19]. In addition, shoulder dysfunction has been reported to contribute to the recurrence and persistence of neck pain [20], suggesting a bidirectional relationship may exist.

In this study, FDR-corrected findings demonstrate a level-dependent pattern of shoulder-cervical morphological relationships; the strongest and most consistent effects were observed at the C5-C6 and C6-C7 levels, particularly for acromion type II and subcoracoid effusion. Higher effect sizes were found to be more common at the mid and lower cervical levels.

Also, our finding that 5.7% of patients with isolated shoulder pain may have findings originating from the cervical spine introduces a clinical consideration. This underscores the potential utility for comprehensive assessments that encompass both shoulder and cervical spine examinations.

Appropriate treatment for shoulder pain associated with the cervical spine may differ from pain related to shoulder pathology. For instance, patients with cervical radiculopathy may benefit from interventions such as physical therapy, medications, and even surgical decompression, while those with shoulder impingement may require a different therapeutic approach [14, 21].

Limitations

While this study provides preliminary insights, several limitations should be acknowledged. First, the cross-sectional design precludes any temporal or causal interpretation of the observed relationships. Longitudinal studies with adequately powered samples, standardized assessments, and appropriate control groups are warranted to validate and extend these findings. Second, since our study was conducted at a single center, it may introduce potential biases related to the specific patient population and healthcare practices in that region. Third, the absence of a control group limits the extent to which findings can be attributed specifically to the study population, as comparisons with individuals without the condition of interest were not possible. Fourth, we did not calculate interrater reliability statistics for the MRI interpretations, which may affect the consistency of subjective assessments such as distinguishing annular bulging from disc herniation. Fifth, the initial presence of sparse data in certain contingency cells led to instability in some effect estimates, reflected by wide confidence intervals. To address this, categories with low event counts were combined based on clinical relevance and interpretability. While this approach improved statistical stability and precision, these findings should be considered hypothesis-generating rather than confirmatory, particularly for variables affected by such restructuring. Sixth, multivariate models are adjusted for age and gender, confounding from unmeasured clinical variables cannot be ignored. Therefore, adjusted estimates should be interpreted as relational rather than causal, and prospective studies with extensive covariate adjustment are needed to determine directionality.

Conclusion

Our study contributes descriptive information on the demographics and associations between cervical and shoulder pathologies. The findings represent morphological associations, between cervical pathologies and shoulder pathologies, and that 5.7% of patients presenting with isolated shoulder pain had cervical pathologies without concurrent shoulder pathology. These findings suggest that cervical spine evaluation may be considered in patients with shoulder pain, particularly in elderly patients. However, given the limitations of this cross-sectional study and the high prevalence of both cervical and shoulder pathology in the general population, these associations should be interpreted cautiously and do not establish causation. Causality cannot be inferred, and the results likely reflect shared degenerative and biomechanical processes rather than direct pathological causation. Prospective studies with control groups and standardized diagnostic criteria are needed to further investigate these relationships.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviation

MRI

Magnetic resonance imaging

PACS

Picture archiving and communication system

PASTA

Partial articular supraspinatus tendon avulsion

ACJ

Acromioclavicular joint

Author contributions

MAA: Conception, design, data collection and processing, literature Review, Analysis; NO: Analysis, data collection and processing, writing; HSK: Data collection, supervision, critical review; SM: Literature review, writing; EA: Data collection and processing; MC: Analysis, data collection and processing; NK: Conception, supervision, critical review. All authors read and approved the final manuscript.

Funding

The authors did not receive support from any organization for the submitted work.

Data availability

The data are not publicly available due to privacy or ethical restrictions.

Declarations

Ethical approval

This study was approved by Clinical Research Ethics Committee of Akdeniz University under the ethical standards laid down in the 1964 Declaration of Helsinki (2020-06-24-461).

Informed consent

Standard clinical informed consent was obtained at the time of the scan. Informed consent for this specific study was not applicable given the retrospective nature of this research.

Competing interest

The authors declare no competing interests.