What muscles should be sampled during Electromyography (EMG) for L2/3, L4, L5, and S1 radiculopathy?

EMG Muscle Sampling for Lumbosacral Radiculopathy

For lumbosacral radiculopathy evaluation, needle EMG should sample specific muscles corresponding to each nerve root level: L2/3 radiculopathy requires iliopsoas and adductor longus; L4 radiculopathy requires vastus medialis and tibialis anterior; L5 radiculopathy requires tibialis anterior, extensor digitorum longus, and gluteus medius; and S1 radiculopathy requires gastrocnemius (medial and lateral heads) and gluteus maximus. 1, 2, 3

L2/3 Radiculopathy Muscle Selection

• Iliopsoas muscle is the primary target for L2/3 root evaluation, as it receives predominant innervation from these levels 1
• Adductor longus should be included as it provides additional L2/3 myotomal representation 1
• These proximal muscles are critical because L2/3 radiculopathies are less common and can be missed if only distal muscles are sampled 3

L4 Radiculopathy Muscle Selection

• Vastus medialis (or medial vastus) is essential for L4 root assessment, as quadriceps weakness is a hallmark of L4 radiculopathy 1, 4
• Tibialis anterior should be sampled as it receives dual innervation from L4 and L5, helping differentiate between these levels 1, 2
• The single leg sit-to-stand test detects quadriceps weakness in 61% of L3/L4 radiculopathies clinically, but EMG provides objective confirmation 4
• Tensor fasciae latae can be added for comprehensive L4 evaluation 1

L5 Radiculopathy Muscle Selection

• Tibialis anterior is a primary L5 indicator muscle and should always be sampled 1, 2
• Extensor digitorum longus (or extensor hallucis longus) provides additional L5 myotomal confirmation 1
• Gluteus medius is valuable for L5 assessment and helps distinguish proximal from distal involvement 3
• L5 radiculopathy shows the highest clinical sensitivity, with hypoesthesia in the L5 dermatome detected in 35.67% of cases, though EMG remains more specific 1

S1 Radiculopathy Muscle Selection

• Gastrocnemius (medial and lateral heads) are the primary S1 muscles, as plantarflexor weakness is characteristic of S1 radiculopathy 1, 2
• Gluteus maximus should be included to assess proximal S1 innervation 1
• Achilles areflexia has the highest positive predictive value (85%) and probability ratio (7.47) for S1 radiculopathy clinically, but EMG provides definitive localization 1

Paraspinal Muscle Sampling

• Paraspinal muscles (multifidus) at the corresponding levels should be sampled in all suspected radiculopathies 2, 3
• Paraspinal EMG increases sensitivity to 100% when combined with limb muscle examination for lumbosacral radiculopathy 2
• However, one study found paraspinal mapping neither sensitive nor specific for exact level localization in a small series, suggesting it should supplement rather than replace limb muscle examination 5
• Sample paraspinal muscles at multiple levels (L3, L4, L5, S1) to identify the specific segmental involvement 1, 2

Critical Technical Considerations

• Multiple sampling sites (2-3 different locations per muscle) are necessary to evaluate recruitment, spontaneous activity, and fibrillation potentials adequately 2, 3
• Needle EMG has sensitivity up to 90% for lumbosacral radiculopathy, making it the most informative neurophysiological method 2
• Nerve conduction studies (NCS) have low diagnostic value for radiculopathy and should not be performed without EMG 2
• EMG should be performed at least 3-4 weeks after symptom onset to allow denervation changes to develop 3

Common Pitfalls to Avoid

• Do not rely solely on distal muscles, as this will miss proximal radiculopathies (L2/3) 1, 3
• Do not skip paraspinal examination, as it significantly increases diagnostic sensitivity 2
• Do not perform EMG too early (before 3 weeks), as acute denervation signs may not yet be present 3
• Clinical examination alone cannot predict EMG results—referred pain has 80.89% sensitivity but low specificity, while specific findings (weakness, areflexia) have high specificity (82.6-97.5%) but low sensitivity (6.37-35.67%) 1