Construct Validity of Joint Position Sense vs. Butterfly Test in Neck Pain Patients
Direct Answer to Your Question
The provided evidence does not contain any information about the "Butterfly Test" for assessing movement sense in patients with neck pain, making a direct comparison of construct validity impossible. However, the evidence does provide substantial data on the construct validity of Joint Position Sense (JPS) testing, also called Joint Position Error (JPE) testing, in this population.
Construct Validity of Joint Position Sense Testing
Evidence Quality and Findings
Joint Position Sense testing demonstrates variable construct validity in patients with neck pain, with significant methodological concerns that limit its clinical utility.
Moderate construct validity was demonstrated in one study where JPE testing successfully discriminated between persons with and without neck pain (AUC = 0.69), though this discrimination was modest 1
Poor construct validity was found in another rigorous study where JPE showed the lowest reliability among six cervical clinical tests (ICC: 0.02 to 0.66) and failed to demonstrate satisfactory construct validity 2
Discriminative validity exists but is limited: JPE testing showed significant differences between chronic neck pain patients and asymptomatic individuals (p < 0.05) in three of four proprioceptive tests, though the head repositioning to 30° rotation test (HR30T) failed to discriminate between groups 3
Reliability Concerns Impact Validity
JPE testing showed moderate to good test-retest reliability (ICC: 0.55 to 0.85) but with high measurement error, making it suitable for group comparisons but of limited utility for individual patient assessments 3
The head repositioning to neutral (HRNT) test performed better than other JPE variants, showing an area under the curve above 0.5 (AUC 95% CI = 0.51-0.78, p ≤ 0.042) and appeared easier to perform in clinical practice 3
Responsiveness as a Validity Indicator
JPE testing demonstrated sufficient internal responsiveness with large effect sizes (Cohen's d range: 0.89-2.00) after 4 weeks of proprioceptive training when measuring absolute JPE 4
However, external responsiveness was inadequate, with weak correlations between JPE measurements and inertial measurement units for most movements 4
Clinical Implications
What This Means for Practice
JPE testing measures a real construct related to cervical proprioception that differs between neck pain patients and controls, but the measurement error is substantial 3, 2
Different JPE tests measure related but dissimilar constructs, with correlations between tests ranging only from 0.35 to 0.61, suggesting they tap into different aspects of proprioception 3
JPE shows minimal correlation with pain-related psychological factors (catastrophizing, fear of movement, disability), with correlations generally below 0.3, indicating it measures a distinct physical impairment rather than psychological constructs 3
Common Pitfalls to Avoid
Do not use JPE for individual clinical decision-making due to high measurement error; reserve it for research or group-level assessments 3
Avoid assuming all JPE tests are equivalent—the head repositioning to neutral test appears superior to other variants for discriminating neck pain patients 3
Do not interpret JPE results in isolation—combine with other sensorimotor tests like the craniocervical flexion test and movement control assessments for comprehensive evaluation 2, 1