The length–force relationship of the sarcomere was originally described by Blix in the late 1800s—a concept that was expanded upon by numerous other investigators [1–5]. In his original experiments, muscle tension during contraction was found to vary as a function of its overall length. After plotting his results, he found that this relationship took the form of a bell curve, the peak of which represented the active tension (i.e., contraction force) produced by a muscle at its resting length. As the length of the muscle was varied, the strength of contraction decreased, even if the muscle was passively stretched prior to initiating contraction (Fig. 3.1). Therefore, suboptimal limb positioning can have a profound effect on muscular contraction strength and, as a result, may lead to inaccurate measurements during the strength evaluation. In other words, maximum contraction strength cannot be achieved when a muscle is placed in a position outside of its native resting length.

Therefore, the reliability of muscular strength testing depends on the clinician’s knowledge of correct testing positions that are designed to optimize the length–force relationship of the particular muscle being tested. A few important clinical examples include strength testing of the infraspinatus, subscapularis, and deltoid muscles as described by Hertel et al. [6, 7]. In these maneuvers, the extremity is passively placed such that the muscle to be tested is in a relatively shortened position and the opposing muscle is in a relatively lengthened position (i.e. increased tension). The patient is then asked to hold the position. If the muscle being tested is weak, its contraction strength cannot overcome the passive tensile force that is applied to the opposing muscle when the arm is released by the examiner. For example, when testing for infraspinatus weakness, the humerus is passively positioned in approximately 20–30° of external rotation (also with the elbow flexed to 90°). This position decreases the passive tension across the infraspinatus muscle since its overall length has been shortened relative to its resting length. On the other hand, this position also increases the passive tension across the opposing subscapularis muscle. When the clinician releases the arm, the patient will attempt to hold the position by contracting the infraspinatus muscle. When this force of contraction cannot overcome the passive tensile force that is applied to the opposing subscapularis, the arm will internally rotate despite the patient’s best efforts (Fig. 3.2). This finding is referred to as a positive external rotation lag sign and the degrees of internal rotation lag (or the amount of internal rotation that occurs after the arm is released) is typically documented as a measure of infraspinatus weakness. The internal rotation lag sign [6], the deltoid lag sign [7], and the teres minor lag sign (also referred to as “Hornblower’s sign” [8] or “drop sign” [6]) use the same length–force relationship concepts and are discussed both later in this chapter and in Chap. 4.

There are numerous muscles that cross or act upon the glenohumeral joint; however, several of these muscles produce similar force vectors which can complicate the assessment of muscular strength. As a result, shoulder motion within any plane likely involves contributions from several different muscles to produce the observed movement. While it would be ideal to isolate and test each individual muscle around the shoulder girdle, complete isolation of a single muscle for the purpose of strength testing is a nearly impossible task. In addition, weakness of one muscle can be substituted by another similarly positioned muscle, masking the underlying weakness during physical examination. This is especially true in cases where subtle weakness is present, such as in small rotator cuff tears where, in many cases, the overlying deltoid may substitute for the deficient cuff. In fact, the ability of the deltoid to substitute for a deficient rotator cuff is the underlying principle of reverse total shoulder arthroplasty in patients with massive, irreparable rotator cuff tears with associated superior migration of the humeral head. Another example includes the levator scapulae and the superior fibers of the trapezius which have similar functions; however, the neural supply to each muscle is different. Thus, a patient with an isolated injury to the dorsal scapular nerve may still demonstrate relatively normal scapulothoracic kinematics despite levator scapulae weakness.

It is typically more advantageous to isolate the overall function of specific groups of muscles rather than attempting to isolate each muscle individually. Alternatively, the clinician can also perform tests that conceptually and theoretically isolate specific muscles with the understanding that complete isolation is probably not attainable. Regardless of the method used, it is necessary to develop a consistent, repeatable examination protocol which can improve individual diagnostic efficiency and accuracy.

Manual muscle testing (MMT) is the most common method by which clinicians evaluate muscle strength. MMT utilizes a standardized grading system that is determined by the ability of the tested muscle act against gravity or against resistance applied by the examiner. In 1916, Lovett and Martin [9] first described the method of manual muscle testing in newborns with infantile paralysis. Since then, abundant research has been conducted regarding its various applications, including modifications of the original grading scale used to describe muscular strength. Despite these modifications, the scale that is most widely accepted is very similar to the original proposal by Lovett and Martin [9] and was devised by the Medical Research Council (MRC) [10] in 1943. The scale has six levels (0–5) and is presented in Table 3.1.

The inter- and intra-observer reliabilities of MMT in the evaluation of various pathologies resulting in muscle weakness range from 0.82 to 0.97 and 0.96 to 0.98, respectively, according to reports dating back to 1954 [11–22]. However, only a few studies have specifically examined the reliability of manual muscle testing for the evaluation of patients with various shoulder pathologies [23–25].

Although the MMT scale is still widely used in clinical practice due to its low cost and rapidity, there are several limitations that must be noted. The first limitation is that the MMT scale is subjective in nature and the score depends on the clinician’s judgment [26–28]. The second limitation of MMT is the inability of the scale to detect small, between-level differences in strength. This is largely due to the stepwise design of the scale and has spurred the development of other scales that have more diagnostic levels [10]. Third, the MMT scale has been criticized for not being capable of detecting clinically relevant differences in muscle strength. MMT was originally developed to measure strength improvements in patients treated with paralytic disorders and muscular dystrophies [29, 30]. Thus, the application of MMT to a variety of clinical settings is probably due to tradition rather than sound scientific rationale. As a result of the subjectivity and reported inaccuracy of MMT, many clinicians (and insurers) prefer to measure strength with more objective means that are more sensitive, such as with handheld dynamometers [31].

A dynamometer is a device used to determine the mechanical force generated by a contracting muscle. While these measurements of force are generally given in Newtons or kilograms, torque can be calculated by simply multiplying Newtons or kilograms by the distance (in meters) between the dynamometer and the center of rotation of the involved joint.

Dynamometers first appeared in 1763 [32] and, since then, numerous modifications have been made. Currently, dynamometers come in a large variety of shapes, sizes, and functional mechanisms that produce the desired force measurements. Isokinetic dynamometers are large machines capable of generating numerous values including peak muscular force, power, and endurance among numerous other measurements (Fig. 3.3) [33]. Isokinetic testing has been used as a standard method of muscle strength measurement over the past 40 years since it has been found to be reliable, reproducible, and valid on numerous occasions [34–38]. As a result, isokinetic devices have also been used as reference standards for the evaluation of newer devices that test muscle strength [39–42].

A large number of studies have evaluated the inter- and intra-rater reliability using handheld dynamometers to assess muscular strength. A systematic review by Stark et al. [43] identified 19 studies in which the authors compared handheld dynamometry to isokinetic muscle strength testing. In that review, all but two studies demonstrated either good to excellent correlation with isokinetic testing or good to excellent intra-class correlation coefficient (ICCs). The study by Burnham et al. [39] found a low correlation of handheld dynamometry with isokinetic testing when measuring shoulder abduction strength in a series of football players (r = 0.28–0.43); however, the scapulae of the tested athletes in that study were not stabilized by the examiner, introducing potential confounding factors in their measurements. Reinking et al. [44] also found a poor correlation between handheld dynamometry and isokinetic testing when measuring knee extension (r = 0.43–0.45); however, testing this group of muscles requires a sufficiently strong examiner to prevent movement of the dynamometer while the subject is tested.

In general, clinical dynamometry is performed with handheld devices due to their portability, simplicity, low cost, and reported excellent reliability and validity when compared to isokinetic dynamometry [27, 43]. Although there are numerous such devices that have been reported as both accurate and reliable for the measurement of muscular force, most handheld dynamometers fall into one of two categories depending on the mechanism of measurement. These include spring scale and strain gauge dynamometers. Spring scale dynamometers work simply by measuring the deformation (lengthening) of a spring as a force is applied—this deformation distance is converted to kilograms and is based on the stiffness (spring constant) of the inserted spring. Strain gauge dynamometers are more complex and work by detecting changes in electrical signals caused by the deformation of an electrical insulator by an outside force (e.g., the force of muscle contraction) (Fig. 3.4).

In 1989, Bohannon and Andrews [45] studied the accuracy of two handheld spring scale and two strain gauge dynamometers using a series of certified weights ranging from 5 to 55 pounds. The dynamometers were tested by gradually increasing the applied weight by 5-pound increments and comparing the readout measurement generated by each device to the actual weight applied. In their study, the spring scale dynamometers measured forces significantly different from the force that was actually applied. The authors also noted that the accuracy of the spring scale dynamometers diminished after extensive use, suggesting that spring fatigue or permanent deformation have been responsible for inaccurate measurements. In contrast, the strain gauge dynamometers measured forces that were much closer to the actual applied force.

Hayes and Zehr [46] evaluated the reliability of MMT, a manual spring scale dynamometer and a digital strain gauge dynamometer to measure rotator cuff strength using a random effects statistical model. In this group of patients with symptomatic rotator cuff disease, they found that the digital strain gauge dynamometer was the most reliable method of measuring rotator cuff strength. Hosking et al. [47] examined the test-retest reliability of handheld dynamometers in children with and without muscular disease and found that repeated testing did not cause measurement variability of more than 15 %. However, another study by Bohannon [28] found the test-retest reliability to be much higher in healthy patients compared to those who had muscle weakness, potentially suggesting that muscle fatigue may play a role in the ability to obtain an accurate measurement of peak muscle force after multiple testing sessions.

There are several other potential limitations of digital handheld dynamometry. The first is that these handheld devices are of minimal use when testing large muscle groups that can produce a much larger force than the examiner can resist. This is particularly true for large, high-output lower extremity muscles that may overcome the strength of the examiner’s upper extremity [28, 48–51]. A second limitation is that an inability to adequately stabilize the device while the subject applies maximal force is quite difficult to achieve. As a result, handheld dynamometers placed in a fixed apparatus have gained popularity to eliminate the effect of examiner strength and stabilization on the reliability of strength measurements [52–55].

Electromyography (EMG) has been used extensively over the past century to evaluate the utility of various manual muscle tests. An electromyogram is obtained by placing an electrode on the skin over the muscle being tested (i.e., surface EMG) or, alternatively, a thin wire can be placed directly into the muscle of interest (i.e., intramuscular EMG) (Fig. 3.5). When the muscle is stimulated, the electrical potential that is produced by the muscle travels through the electrode and towards the connected electromyograph which interprets and displays the signal through an oscilloscope. It is important to remember that EMG readouts with higher amplitude do not necessarily indicate that the muscle is generating greater force. As an example, an eccentrically contracting muscle produces similar amplitude as a concentrically contracting muscle; however, the force produced by the eccentric contraction may be much less than that produced by the concentric contraction.

EMG is an important tool for the evaluation of skeletal muscle activity; however, its interpretation can be influenced by several factors that must be taken into account. Features of the surface electrode such as width, diameter, and electrical properties can influence the signal output. In the case of surface EMG, increased distance or increased soft-tissue interposition between the surface electrode and the muscle being tested can also significantly influence signal interpretation [56, 57]. The primary drawback of thin-wire EMG is that the sample size is limited to the surface area of the small electrode whereas surface EMG can obtain measurements over an expanded area of muscle tissue and is also easier to implement; however, this can also introduce unwanted noise due to soft-tissue interposition and contributions from surrounding musculature. In addition, the amplitude or morphology of the EMG readout may be affected by the type of muscle being tested (fast-twitch versus slow-twitch).

Many studies have utilized a normalization technique to study muscle activity—that is, the electrical measurements are compared to a reference standard generated from a maximal voluntary contraction (MVC) of the muscle in question. The ratio of the two measurements is recorded and compared between different subjects [58]. Other methods of obtaining EMGs involve submaximal voluntary contractions and isometric measurements; however, these methods have been less reliable to date [57, 59].