Lower trapezius muscle function in people with and without shoulder and neck pain: a systematic review

Search strategy

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. Ovid (MEDLINE), PubMed, and ScienceDirect were searched from inception until July 20, 2021. No language restrictions were applied to conference abstracts or published articles. In order to perform an all-encompassing and broad search, these keywords were utilized: (Lower Trapezius) AND (Pain OR Posture OR Dysfunction).

Eligibility criteria

Case-control studies including participants of any age with upper body pain of any type (specific or nonspecific) and of any duration (acute, subacute, or chronic) were eligible for inclusion. Studies measuring lower trapezius function in any way were included. For example, strength, fatigability, percentage of maximum voluntary contraction, timing or extent of contraction, cross-sectional area, or other unidentified measurement were eligible for inclusion. Studies could assess the lower trapezius with any clinical measurement tool or quantifiable assessment, with the participant in dynamic or static activity. Studies were excluded if the subjects were solely determining the effect of an intervention on the lower trapezius muscle.

Study selection

One reviewer conducted the electronic searches (DW). Two other reviewers (NH/CL) independently screened citations at the title and abstract levels. Of the total studies identified, each citation was reviewed to identify lower trapezius effects by two independent reviewers. Data were independently extracted by two reviewers (NH/CL). Seven of the studies led to confusion whether to be included, so a third reviewer (DW) cross-checked, and a conclusion was made to include four and exclude three studies. Due to the heterogeneity among studies, a meta-analysis was not carried out.

Patient and public involvement

There was no patient and public involvement in the production of this research.

Study identification

Searches retrieved 1,274 citations, of which 110 were eligible for full-text review. After review, 18 full-text articles of mixed methodological quality were included, whereas 92 were excluded based on the exclusion criteria and not being able to retrieve reports (Figure 1). Among the 92 excluded articles, nine of them could not be retrieved because they were not provided by authors. Thus, they were subsequently excluded. This can be seen in Figure 1.

Figure 1:

PRISMA flow diagram.

Of the 18 studies that were included, two-thirds of the studies (66.7%) indicated that the lower trapezius muscle had a significant impact on neck and shoulder pain [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40]. Within the studies, there was an even split between the lower trapezius muscle having an impact on shoulder pain (6/12) [25, 26, 28, 29, 31, 34] vs neck pain (6/12) [23], [24], [25], [26], [27, 30, 32, 33]. Within the shoulder pain group, half the studies involved the lower trapezius having a significant impact (6/12) [25, 26, 28, 29, 31, 34] vs a nonsignificant impact (6/12) [35], [36], [37], [38], [39], [40]. Interestingly, in the neck pain group, the lower trapezius muscle had a 100% impact [24, 27, 30, 32, 33]. These descriptions can be visualized in Table 1 and Figure 2.

Figure 2:

The impact of the lower trapezius on shoulder and neck pain in the included studies.

Characteristics of included studies

The 18 articles investigating lateral trapezius function included 431 participants with pain and 405 healthy controls without pain [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40]. The studies included one or more of the following lower trapezius measurement outcomes: electromyographic (EMG) activity level [25, 28, 32, 33, 35, 36, 39], EMG time to onset [24, 26, 31, 33, 34, 36, 39], EMG time of peak activation [24, 38], lower trapezius strength [23, 27, 29, 40], or muscle thickness [30, 37]. These details are further elaborated in Table 2.

Included studies by measurement outcome

The included studies per measurement outcome are summarized in the following sections and in Table 3.

Level of muscle activity (EMG)

Seven studies analyzed lower trapezius muscle activity with EMG [25, 28, 32, 33, 35, 36, 39]. All seven studies measured lower trapezius muscle activity dynamically, with mixed results. Three studies observed subjects with shoulder pain during functional tasks [25, 35, 36]. Specifically, Sabzehparvar et al. [35] studied 24 elite male swimmers, ages 15–21, with and without mechanical pain, and found maximal voluntary contraction (%MVC) values of 10.9% (±8.12%) and 9.3% (±4.95%), respectively (p=6.09). Meghdadi et al. [36] studied 60 elite male table tennis athletes, ages 23–27, with and without shoulder impingement syndrome (SIS), and reported muscle activity values of 37 (±5.0) and 39.2 (±6.3), respectively (p=0.301). Shih et al. [25] studied shoulder kinematics in 20 patients with frozen shoulder against 20 healthy controls, in the age range of 45–60, and found a 55.9% (p=0.001) decrease in lower trapezius muscle activity during the scaption task compared to controls.

Two studies analyzed the effects of neck pain on lower trapezius muscle activity [32, 33]. Zakharova-Luneva et al. [32] studied 18 neck pain patients against 20 healthy controls, ages 18–34, and reported significant increases in lower trapezius muscle activity during shoulder abduction and external rotation compared to controls (p<0.025, p<0.034, respectively). Helgadottir et al. [33] analyzed upper limb elevation in 22 patients with insidious onset neck pain (IONP) against 23 asymptomatic controls, ages 22–43. They found right lower trapezius muscle activity in IONP to be insignificantly greater than that of controls (0.84 [0.05]) and 0.82 [0.04], respectively, p values not reported) and left lower trapezius muscle activity in IONP to be insignificantly lower than that of controls (0.78 [0.05] and 0.80 [0.04], respectively, p values not reported) [33].

The remaining two analyzed muscle activity under load-bearing conditions [28, 39]. Michener et al. [28] observed a statistically significant decrease in lower trapezius muscle activity during active arm elevation with weights adjusted to participant BMI in 28 subjects (ages 24–52) with subacromial pain syndrome when relative to upper trapezius and serratus anterior (SA) muscle activities. Specifically, they compared these subjects to 28 controls (ages 24–52) and found a higher upper trapezius/lower trapezius ratio in the pain group during ascending (mean difference=0.92, p=0.008) and descending (mean difference=0.70, p=0.03) movements and a lower trapezius/serratus anterior ratio in the pain group during ascending (mean difference=−0.25, p=0.026) and descending (mean difference=−0.51, p=0.032) movements [28]. Larsen et al. [39] observed a nonsignificant higher level of lower trapezius muscle activity (p=0.11, raw data not reported) during weighted shoulder elevation in 16 subjects (ages 27–55) with subacromial impingement syndrome against 15 healthy controls (ages 27–51).

Muscle activation (EMG)

Two studies analyzed lower trapezius muscle activation [24, 38]. Ghaderi et al. [24] studied 40 individuals with and without shoulder pain (ages 22–39) to observe differences in time to reach peak activation between the two groups, but their conclusions were statistically insignificant (1.91 ms [0.55] and 1.89 ms [0.64], respectively, p=0.97]. Larsen et al. [38] analyzed the ability to selectively activate the lower trapezius muscle through EMG and observed that fewer subjects with SIS were able to successfully selectively activate the lower compartments of the trapezius as defined by activation ratios equal to or greater than 95.0% when compared to their healthy controls. Specifically, out of 15 patients with SIS (ages 27–53), three were able to reach selective activation (p=0.03) compared to the 9 out of 15 controls (ages 27–51) who attained selective activation [38].

Time to onset (EMG)

Seven studies reviewed the amount of time it took for the lower trapezius to activate [24, 26, 31, 33, 34, 36, 39]. Cools et al. [34] and Phadke and Ludewig [31] both studied subjects with SIS, evaluating trapezius muscle latency. Cools et al. [34] compared 39 subjects with SIS (ages 20–32) against 30 healthy controls (ages 18–26) and found that the lower trapezius on the injured side of individuals with SIS had a longer mean muscle latency time of 174.3 ms (38.9), which is 31.5 ms longer (p<0.01) than the mean muscle latency time of the dominant side of the controls, which was reported to be 142.8 ms (33.1). On the contrary, Phadke and Ludewig [31] studied 24 patients with SIS (ages 23–47) against 25 healthy controls (ages 22–42) to determine the lower trapezius latency relative to the serratus anterior and upper trapezius during unloaded, loaded, and after repetitive motion conditions. They found that, relative to the serratus anterior, the lower trapezius in healthy controls had a higher latency than that in patients with impingement under unloaded conditions, but they found the opposite result under loaded and after repetitive motion conditions (results were graphically reported, raw data unavailable) [31].

Leong et al. [26] studied 43 male volleyball players with and without rotator cuff tendinopathy (RCT) (ages 18–27) and found that the lower trapezius of the RCT group is activated slower than that of the control group (−18.5 ms [43.9] vs. −72.4 ms [39.0], respectively; p=0.001). Helgadottir et al. [33] studied the time to onset of lower trapezius muscle activation on both sides, and concluded that there is an insignificant increased delay to onset in the 22 IONP patients compared to the 23 healthy controls (right side: 0.43 s [0.10] IONP vs. 0.34 s [0.07], respectively; left side: 0.48 s [0.08] vs. 0.36 s [0.05], respectively). Similarly, in the 20 patients with neck pain in the study by Ghaderi et al. [24], they found an insignificant delay in lower trapezius muscle onset time (0.21 ms [0.21] vs. 0.14 ms [0.15]) compared to the 20 controls (p=0.18). On the contrary, Meghdadi et al. [36] observed an insignificant reduction in time to onset in the 30 subjects with pain compared to their 30 healthy controls (−386.93 ms [31.02] and −384.12 ms [31.09], respectively; p=0.807). Finally, Larsen et al. [39] studied lower trapezius muscle onset time in 16 cases of SIS against 15 control cases under nonloading and load-bearing conditions, and they observed an insignificant difference in time to onset in subjects with pain in these muscle pairs: upper trapezius/lower trapezius (p=0.98) and lower trapezius/serratus anterior (p=0.53).

Muscle strength

Four studies utilized handheld dynamometers to consider strength in the lower trapezius [23, 27, 29, 40]. Three of these studies were in agreement and observed a significant decrease in lower trapezius muscle strength in subjects with pain compared to their healthy controls. Specifically, of these three studies, Park et al. [23] and Petersen et al. [27] both measured muscle strength in subjects with unilateral neck pain and found significantly reduced muscle strength in the lower trapezius ipsilateral compared to that contralateral to the pain. Park et al. [23] observed 26 female violinists (ages 20–22) with unilateral neck pain, studying their contralateral, nonpain side as controls, and found lower trapezius strength on the side with pain to be 26.9 N (13.10), whereas strength on the contralateral side was 29.8 N (15.8), p=0.02. Similarly, Peterson et al. [27] observed 22 subjects with idiopathic neck pain (ages 19–45) with 17 healthy controls (ages 20–44), first analyzing the strength of the lower trapezius muscles ipsilateral and contralateral to the site of neck pain in the pain group and the strength of the right and left lower trapezius in the controls, and then comparing within groups and between groups. They found that in the pain group, the ipsilateral lower trapezius had a measured strength of 18.9 N (6.5), whereas the contralateral lower trapezius had a measured strength of 21.5 N (9.4) (p<0.01) [27]. Additionally, the pain side adjacent in the control group—the right side—had a measured muscle strength of 24.4 N (8.5), which is significantly greater than the ipsilateral lower trapezius in the pain group (p=0.02) [27]. Leong et al. [29] analyzed 66 athletes with and without RCT (ages 18–28) and also observed a decrease in lower trapezius strength by 24.3% in those with RCT compared to healthy controls (p<0.05). The remaining study by Trakis et al. [40] observed no significant changes in lower trapezius muscle strength in 12 male baseball pitchers, ages 15–17, with shoulder and elbow pain compared to the 11 healthy male controls, ages 15–17 (strength imbalance of 9% [17] and 11% [16], respectively; p=0.74).

Muscle thickness

Two studies analyzed lower trapezius muscle thickness through ultrasound imaging [30, 37]. McKenna et al. [37] measured lower trapezius thickness at rest and submaximal contraction between 52 swimmers (ages 34–62) with and without shoulder pain. At rest, the lower trapezius of those with pain had a thickness of 5.1 mm (4.5–6.7), while the lower trapezius of the controls had a thickness of 4.6 mm (4.4–6.2), although this difference was insignificant (p=0.59) [37]. During contraction, they found the lower trapezius of those with pain to have a thickness of 8.8 mm (2.0), while the lower trapezius of the controls had a thickness of 9.3 mm (2.3), although this difference was insignificant (p=0.36) as well [37]. Similarly, Uthaikhup et al. [30] measured lower trapezius muscle thickness at rest at 0° abduction and at rest at 120° abduction in 40 female subjects with and without right-side neck pain (ages 21–32), but instead found a significant decrease in muscle thickness in neck pain subjects compared to controls at 0° rest (2.39 mm [0.68] and 3.04 mm [0.66], respectively [p<0.01]) and 120° rest (2.83 mm [0.96], 3.47 mm [0.78], respectively [p<0.05]), as well as a significant decrease in lower trapezius thickness from rest to contraction in pain subjects compared to healthy controls (3.81 mm [1.96], 5.27 mm [1.63], respectively [p<0.05]).

Limitations

This systematic review was created to be comprehensive and thorough. However, it may be possible that not all appropriate studies were included. The chances of this happening were reduced by an all-encompassing search methodology, individual abstract and title screening by two researchers (NH/CL), and strict adherence to PRISMA guidelines. Further, limitations such as only a small number of studies per measurement outcome, unclear definitions of the duration and type of shoulder and neck pain, and variation in study methodology have precluded more advanced methods of analysis. This variation may play a part in explaining some insignificance between controls and comparators within studies. They could have additionally diluted this systematic review’s results, which may explain why some of these findings are unclear for some measurable outcomes. As in many studies, differences in reliability of measures may also affect outcomes of the included studies, especially if they may be based on how well versed a practitioner is at utilizing equipment that relies on it. Further, only 11 of the included studies reported measurement reliability, with an increased variability among studies for the indicated measurement outcomes [24], [25], [26], [27], [28], [29], [30], [31, 35, 37, 40]. If reliability is poor, this can influence the findings and cause more insignificance in which the differences between the comparator and control groups are relatively small. As such, this may have an impact on the results of the included studies in this systematic review. A thorough investigation should be conducted on the existing reliability in this field. The significance and results of this finding should be taken with caution given the existing limitations of this review and the individual studies. However, it provides a comprehensive summary of the literature that can be of use to clinical practitioners and researchers.