Lower-Extremity Muscle Cross-Sectional Area After Incomplete Spinal Cord Injury

Methods 

Participants 

People with incomplete SCI 

A convenient sample of 17 people (15 men, 2 women; time postinjury, 13±9mo) with incomplete SCI (age, 39±15y; body mass, 76±13kg; height, 178±10cm) volunteered to participate in the study. Of these, 10 subjects were studied at the University of Florida, Gainesville, FL, and 7 subjects were studied at the University of Georgia, Athens, GA. All participants (1) had a diagnosis of traumatic SCI at cervical or thoracic levels (C4-T12) resulting in upper motoneuron lesions in the lower extremity, (2) had a history of grade C or D SCI as defined by the American Spinal Injury Association (ASIA) Impairment Scale, and (3) had a medically stable condition at the time of testing. Seven people with incomplete SCI used a wheelchair, 2 used forearm crutches, and 6 used a single-point cane for community ambulation (table 1).

Table 1. Characteristics of Subjects After Incomplete SCI

	Subject No.	Time Postinjury (mo)	Injury Level	ASIA Grade	Mobility Status	Orthosis	LEMS
	S1	5	C5	D	WC	NA	35
	S2	10	C5	D	Forearm crutch	AFO	40
	S3	13	C5	D	Single-point cane	AFO	43
	S4	15	C6	C	WC	NA	35
	S5	16	C5	C	WC	Bil-AFO	33
	S6	20	T1	D	Single-point cane	NA	44
	S7	37	T1	D	Bilateral forearm crutches	AFO	35
	S8	15	T1	D	WC	NA	45
	S9	11	C4	C	WC	NA	17
	S10	18	C6	D	Single-point cane	NA	NT
	S11	7	C6	C	WC	NA	26
	S12	13	T4	C	WC	NA	20
	S13	7	C6	D	Single-point cane	AFO	39
	S14	7	C3	D	Single-point cane	AFO	49
	S15	7	T7	D	Single-point cane	NA	43
	S16	12	T12	D	No assistive aid	NA	45
	S17	12	C5	D	No assistive aid	NA	49

Abbreviations: AFO, ankle-foot orthosis; Bil, bilateral; LEMS, lower-extremity muscle score (normative, 50/50) as assessed by the ASIA motor impairment scale34; NA, not applicable; NT, not tested.

Controls 

Seventeen people (15 men, 2 women; age, 39±12y; body mass, 78±12kg; height, 178±8cm) volunteered to serve as control subjects. These subjects were matched to incomplete SCI subjects on the basis of age, sex, height, and body mass. Although large demographic variability existed among subjects in both the incomplete SCI and control groups, each control person was closely matched (age, ±7y; height, ±10cm; body mass, ±8kg) to the corresponding person with incomplete SCI. Control subjects were recreationally active but not engaged in any rigorous exercise program and were recruited from the Gainesville community.

Before participation in the study, all subjects were informed of the purpose of the investigation and all provided written informed consent as approved by the institutional review boards at their respective universities. The conduct of all investigation conformed to the protocol and the ethical and humane principles of research.

Maximum Muscle CSA 

Proton magnetic resonance imaging (MRI) was used to determine maximum muscle CSA of the lower extremity. MRI for all subjects was performed specifically for the study. A 1.5-T superconducting magnet scannera was used to collect transaxial images of the leg and thigh. A standard (length, 20cm) lower-extremity quadrature coil or body coil was used for imaging. A fast gradient-echo or spin-echo imaging sequence was used with the following imaging parameters: acquisition matrix size of 256×256 to 256×192 pixels, field of view of 16 to 32cm for the leg and 22 to 40cm for the thigh, pulse repetition time of 51 to 300ms, echo time of 10 to 27ms, slice thickness of 5 to 7mm, and slice gap of 0 to 5mm. Subsequently, the fat-free maximal muscle CSA of lower-extremity muscles was determined by using a custom-designed interactive computer program as previously described.21 Specifically, the maximum CSAs of the quadriceps femoris (QF) and hamstring (HAMS) muscle groups in the thigh; the soleus (SOL), medial gastrocnemius (MG), lateral gastrocnemius (LG), and tibialis anterior (TA) muscles in the lower leg; and the entire posterior compartment (PC) of the lower leg (including the tibialis posterior [TP] muscle) were calculated (fig 1). In addition, maximum CSA of the ankle antigravity muscles (plantarflexors [PFs]) was considered at the level of the lower leg where the soleus, medial gastrocnemius, and lateral gastrocnemius taken together resulted in the largest CSA.

View Large Image Download to PowerPoint

Fig 1. Representative transaxial proton MRIs of (A, B) the leg and (C, D) thigh in a subject with incomplete SCI (right) and corresponding age-matched control (left). Abbreviations: A, anterior; L, lateral; M, medial; P, posterior.

Results 

Maximum Muscle CSA After Incomplete SCI 

Lower-extremity muscle size in pooled incomplete SCI subjects was significantly smaller than that of the control group for all tested muscles (P≤.004). Mean differences in muscle CSA ranged from 24% (tibialis anterior) to 31% (quadriceps femoris) in the incomplete SCI group relative to the control group. Muscle-specific CSA data are presented in figure 2.

View Large Image Download to PowerPoint

Fig 2. Muscle CSA in the pooled incomplete SCI and control groups of (A) leg muscles and (B) thigh muscles. NOTE. Values are means ± standard error of the mean. *Significant differences between incomplete SCI and control subjects.

Impact of Ambulatory Status on Maximum Muscle CSA After Incomplete SCI 

Our results showed that people with incomplete SCI in both the WC and non-WC groups had a smaller skeletal muscle size compared with controls. Subjects in the WC group had significantly smaller muscle CSA values for all of the antigravity muscles (ie, soleus, medial gastrocnemius, lateral gastrocnemius, quadriceps femoris) relative to their corresponding control group (P≤.048) (table 2). Mean differences in lower-extremity–muscle CSA in the WC group ranged from 21% (quadriceps femoris, tibialis anterior) to 39% (medial gastrocnemius). Neither the tibialis anterior (P=.064) nor the hamstring (P=.120) differed significantly in the WC group compared with matched controls. The non-WC incomplete SCI group also showed significant differences in muscle CSA values relative to control subjects. With the exception of the medial gastrocnemius, all of their lower-extremity–muscle CSA values were significantly smaller than those measured in the control group (P≤.021). Mean differences in muscle CSA in the non-WC group ranged between 24% (soleus) and 38% (quadriceps femoris) relative to the control group (see table 2).

Table 2. Percentage Differences Between the Lower-Extremity Maximum Muscle CSAs of the WC and Non-WC Groups Relative to Corresponding Controls

	Muscle Groups	Individual Muscles	% Difference
			WC Group (n=7)	Non-WC Group (n=10)
		Soleus	27.5±7.5⁎	23.8±8.7†
	Plantarflexors	Medialgastrocnemius	38.8±5.3⁎	13.7±11.8
		Lateralgastrocnemius	30.1±9.8⁎	24.7±10.3†
	Dorsiflexors	Tibialisanterior	20.7±6.1	26.1±8.3†
	Knee extensors	Quadricepsfemoris	21.4±3.7⁎(n=6)	37.5±9.3†(n=8)
	Knee flexors	Hamstring	23.1±5.1(n=6)	34.9±6.8†(n=8)

NOTE. Values are differences in percentage means (% difference) ± standard error of percentage.

⁎ Significant difference between the WC and control groups (P<.05). † Significant difference between the non-WC and control groups.

Differential Atrophy After Incomplete SCI 

Comparisons in the magnitude of atrophy across affected lower-extremity muscles showed no differential atrophy between extensor and flexor muscle groups about the ankle (PF:TA, P=.433) or knee (QF:HAMS, P=.769) in the pooled incomplete SCI group and control groups (table 3). However, similar comparisons suggest greater relative atrophy of the antigravity muscles in the leg in the WC group compared with the non-WC group (PF:TA, P=.043) (fig 3). When examined separately, the proportion of the overall plantarflexor CSA occupied by soleus or lateral gastrocnemius did not differ between the pooled incomplete SCI and control groups (see table 3). However, the proportion of overall CSA occupied by the medial gastrocnemius was significantly smaller in the WC group than in the non-WC group (MG:PC, P=.002) (fig 4). Finally, the relative ratios between proximal and distal antigravity muscle CSA (QF:PF) were similar for all comparisons made (P≥.273).

Table 3. Relative Proportions of Muscles in the Pooled Incomplete SCI and Control Groups

	Description	Proportions	Incomplete SCI	Controls
	Individual PFs relative to PC	SOL:PC MG:PC LG:PC	0.47 0.27 0.16	0.47 0.27 0.17
	Intracompartment ratios⁎	PF:TA	5.08	5.53
		QF:HAMS	1.94	1.99
	Intercompartment ratios†	QF:PF	1.74	1.70

Abbreviations: HAMS, hamstring; LG, lateral gastrocnemius; MG, medial gastrocnemius; PC, posterior compartment; PF, plantarflexor; QF, quadriceps femoris; SOL, soleus; TA, tibialis anterior.

⁎ Flexor and extensor ratios of leg and thigh. † Proximal-distal antigravity muscle ratios.

View Large Image Download to PowerPoint

Fig 3. Proportion ratios of muscle groups within the leg and thigh (PF:TA, QF:HAMS) and proximal and distal antigravity muscles (QF:PF) between the WC and non-WC groups. *Significant differences between WC and non-WC incomplete SCI subjects. Abbreviations: see table 3.

View Large Image Download to PowerPoint

Fig 4. Ratio of individual plantarflexor muscles to the maximum CSA of the PC of the leg (SOL:PC, MG;PC, LG:PC) in the WC and the non-WC group. *Significant differences between WC and non-WC group incomplete SCI subjects. Abbreviations: see table 3.

Discussion 

The findings of our study indicate marked atrophy of lower-extremity muscles after incomplete SCI. Overall, subjects with incomplete SCI had a 24% to 31% smaller average muscle CSA in affected lower-extremity muscles compared with control subjects. Mean differences were highest in the thigh muscles (≈31%) compared with the leg muscles (≈25%), with no differences noted between the self-reported more- and less-involved limbs within the incomplete SCI group. Dichotomizing the incomplete SCI group into WC and non-WC users showed that both the WC and non-WC groups had a significantly lower muscle CSA than their respective control groups. In addition, the WC group exhibited greater atrophy in the ankle plantarflexor muscles compared with the dorsiflexors, suggesting a preferential atrophic response in the antigravity muscles.

The magnitude of atrophy in our subjects after incomplete SCI is much less than that reported in people after complete SCI. An overall 46% decline in average CSA of the lower-extremity muscles in people after complete SCI has been reported 24 weeks after injury, with decreases in the soleus (68%), gastrocnemius (54%), tibialis anterior (20%), quadriceps femoris (42%), and hamstring (44%) muscles reported relative to controls.23 These values (except for the tibialis anterior muscle) are approximately twice those seen in the present study. One of the main reasons is likely related to the partial sparing of voluntary motor control after motor incomplete SCI. Skeletal muscle atrophy after SCI is a result of primary injury to motoneurons in the spinal cord and concurrent inactivation of affected skeletal muscle along with subsequent changes in muscle length and mechanical loading conditions.24 Fractional presence of neural inputs to the affected muscle allows for variable activation of lower-extremity musculature after incomplete SCI. In fact, research subjects in our study were typically able to load their lower extremities during transfers and/or ambulate with crutches or a cane.

An interesting finding in the present study was the significant difference in lower-extremity muscle CSA in both the WC and the non-WC incomplete SCI groups. When compared with the control group, the WC group showed atrophy in all the antigravity muscles. The non-WC subjects showed a similar pattern, with the exception that no differences were seen in the medial gastrocnemius. One likely explanation for the atrophy seen in both incomplete SCI groups is that people who use assistive aids (eg, cane, crutches) for ambulation transfer much of the weight-bearing demands of daily activities to their upper extremities.25, 26 Consequently, weight-bearing through the affected lower extremities is reduced. This reduced use and activation of the affected lower-extremity muscles triggers a cycle of added muscle atrophy and further dependence on assistive devices. In a recent study, Clark et al27 showed reductions in leg muscular activity after use of a crutch or walker for ambulation in able-bodied subjects. The researchers found that the knee extensor (vastus lateralis) and ankle plantarflexor muscles (soleus) incurred maximum decreases in muscle activity, thereby suggesting a predisposition of the antigravity muscles to dysfunction after unloading. This finding is consistent with both animal and human models of unweighting.28, 29

Unlike antigravity muscles the proximal and distal flexors in the WC and non-WC groups showed varied findings. The ankle dorsiflexor muscles showed significant differences in CSA in the non-WC group (≈26%) but not in the WC group (≈21%). This finding seems counterintuitive given the perceived differences in the pattern of muscle activation between WC and non-WC users. However, it has been reported that antigravity extensor muscles contribute much more than the flexors to an efficient gait.30, 31, 32 Therefore, the discrepancy noted in the ankle and knee flexor muscle response between the 2 groups is probably not related to activation patterns during gait. Unlike the WC group (1/7), a greater proportion of the subjects in the non-WC group (5/10) were fitted with ankle-foot orthoses for daily use (see table 1). This characteristic may serve to explain the greater magnitude of atrophic response in the non-WC group given the limited dorsiflexion range of motion allowed with an orthosis. However, the use of an orthosis did not necessarily correspond with lower-extremity motor scores (see table 1) in our incomplete SCI groups. As such, we cannot fully explain the dorsiflexor atrophic response in the non-WC group.

When the pooled subject data were examined, the proportion of individual plantarflexor muscles relative to the compartments they occupy was consistent between the incomplete SCI and control groups. This finding suggests that all the muscles studied underwent similar relative amounts of atrophy, independent of mobility status. However, when the WC and non-WC groups were examined individually, the medial gastrocnemius muscle relative to the posterior compartment (PC) in the WC group showed a differential atrophy compared with the other groups. Similarly, although no differences existed between the plantarflexor-to-tibialis anterior ratios in the pooled incomplete SCI and control groups, significantly lower plantarflexor-to-tibialis anterior ratios are noted in the WC group versus the non-WC group, suggesting more overall atrophy of the plantarflexors in the WC group (≈32% vs 21% in the WC vs non-WC, respectively). The most intuitive explanation for this finding would be the greater relative loading imposed on this muscle group during an upright (nonwheelchair) versus a seated posture (wheelchair). Previous studies have documented that the ankle plantarflexors are critical during locomotion and generate most of the propulsive forces necessary for efficient walking.31, 32 Because walking is comparatively more compromised in people who use a wheelchair versus a cane or crutch for ambulation, mechanical loading via weight-bearing on the paralyzed muscles is less. In addition to differential loading, a prolonged flexed position of the knee during wheelchair use might shorten the gastrocnemius muscle at the knee joint, resulting in greater plantarflexor muscle atrophy relative to the non-WC group.

In contrast to differences in leg muscle atrophy, differences in the thigh muscle CSA in the WC and the non-WC groups were not significant. The quadriceps femoris-to-hamstring ratio was similar in both incomplete SCI groups, implying relatively similar atrophy of the thigh musculature independent of ambulatory status. Last, the degree of atrophy between the antigravity proximal and distal extensor muscles (QF:PF) was also similar. Collectively, our results suggest that although using a cane or crutch for ambulation might attenuate the atrophic response of the ankle plantarflexor muscles (primarily the medial gastrocnemius), marked adaptations are seen throughout the entire lower extremity after incomplete SCI.

The atrophic response in people after incomplete SCI as described in this study most likely has dramatic functional implications. Previous studies12, 13, 14 have shown that decreases in muscle CSA are strongly related to impaired muscle strength; in addition, muscle strength after incomplete SCI plays an important role in functional walking performance.33 Also, gains in skeletal muscle size have been associated with improvements in motor function. For example, in a study9 assessing motor and sensory recovery after incomplete SCI, hypertrophy of the partially innervated skeletal muscles was suggested as a factor that could account for the motor recovery after rehabilitation. As such, if specific deficits associated with incomplete SCI can be discerned, efficient strategies can be designed for functional rehabilitation.

There are some shortcomings in our study. Subjects with incomplete SCI were, on an average, 13 months postinjury (range, 5–37mo). Motor recovery after incomplete SCI is a continuous process that reaches a plateau around 1 year postinjury, with a significantly slower rate of recovery in the second half-year interval after injury.9 As a result, subjects with incomplete SCI in this study were at slightly different stages of motor recovery and not necessarily in a steady condition.9 Furthermore, our subjects underwent routine rehabilitation treatment before participation in the study. Therefore, it is possible that the true values of actual atrophy might have been underestimated. In addition, we cannot definitely confirm the subjects’ preinjury CSA values to be similar to those of the control group. However, our control subjects were matched in sex, age, height, and weight and have comparable CSA values to controls in other human studies.23 Despite these limitations, this study provides unique findings regarding the impact of ambulatory status on muscle CSA in people with incomplete SCI.

Conclusions 

This study shows that incomplete SCI is associated with significant muscle atrophy in the affected lower extremity that is uniform between limbs and is somewhat influenced by mobility status. Most therapeutic approaches for improving locomotor performance in subjects with incomplete SCI are compensatory rather than physiologically based. As such, people after incomplete SCI are often left with significant motor deficits despite long-term therapeutic intervention. With the increasing prevalence of people living with incomplete SCI, there is an urgent need to develop appropriate therapeutic techniques with the goal of maximizing motor recovery, thereby reducing disability. When developing therapeutic interventions to enhance functional recovery after SCI, an understanding of the underlying physiology of muscular responses that occur after this type of injury may promote different intervention strategies that rely less heavily on compensatory rehabilitation. This study will provide a foundation from which the relation between muscle size and function in this population can be further explored. Future research efforts can be directed toward understanding relations between physiologic deficits in skeletal muscle after SCI and parameters of functional ability like walking balance, speed, and muscle strength.