Inclusion criteria were diagnosis of cerebral palsy with spasticity as a predominate feature, prior diagnosis by a pediatric neurologist, less than 12 years of age, and completion of baseline assessment and informed consent forms. Exclusion criteria were recent interventions, such as injections with botulinum toxin type A within 4 months of study enrollment. As noted, all individuals for whom baseline assessments were completed and informed consent were received were included in the study regardless of their participation in and completion of study protocols in the larger OMT versus acupuncture investigation. Approximately 40 subjects completed both studies. Approval for this study was obtained from the institutional review board at the University of Arizona in Tucson.
Data were taken from the baseline assessments in the larger study. Children were recruited from local clinics, where the diagnosis of spastic cerebral palsy was made by a pediatric neurologist.
16 On study enrollment,
16-18 subjects received an extensive battery of measures designed to evaluate impairment, functioning, and well being. The parent and, where appropriate, the child signed informed consent forms.
Osteopathic objective physical findings (ie, somatic dysfunction) were collected and included:
Although osteopathic examiners process the results of musculoskeletal structural examinations somewhat differently as a result of their own physiology, training, and clinical experience, we sought to reduce the effects of this bias by having only one osteopathic physician (K.W.) complete all baseline subject assessments. The two variables analyzed in the current study were fascial restrictions and restriction of spinal motion. Findings were recorded on an original form (Appendix) that included a 100 mm visual analog scale to be used by the examiner to rate the severity of subject muscle spasticity.
Validity refers to how adequately a test or instrument measures what it purports to measure. There are several types of validity including construct validity and predictive validity.
Construct validity refers to the degree to which an underlying hypothesized dimension is represented or confirmed by empirical data.
Concurrent validity refers to the relationship of a measure to an external criterion.
19 (Alternative terms for these two concepts are
factorial validity and
predictive validity, respectively.) In this study, construct validity was used to assess the structure and interrelationships of the osteopathic findings. Concurrent validity was assessed by the correlation between several osteopathic latent factors and an external criterion—a rating of the child's muscle spasticity on a visual analog scale.
We specified latent factors for fascial restrictions and restriction of motion in the spine. As noted, we hypothesized that these two factors would correlate with muscle spasticity. Children with greater restrictions in these factors were expected to have higher levels of impairment as reflected by examiner ratings of muscle spasticity.
The fascia are sheets of connective tissue that wrap around every muscle group and internal organ and are continuous and contiguous from head to toe. The four primary fascia in the transverse plane are the tentorium cerebelli, cervicothoracic junction, abdominal diaphragm, and the pelvic diaphragm. Blood, lymphatic vessels, and nerve fibers pierce through and are held in place by the fascia. According to osteopathic theory, restrictions or twists in the fascia can have localized or distant effects on tissue health by changing local levels of oxygenation, nutrition, detoxification, and neuronal activity. The four fascial planes act as transverse baffles in the closed-fluid system of the human body. If they move in synchrony, a fluid wave can pass through the system with minimal disturbance; when out of synchronization, the same fluid wave produces complex patterns of disturbance. This action, in turn, disturbs the milieu of the cell and the homeostasis of the organism, affecting basic functionality. The osteopathic structural examination assesses the presence and severity of restriction in the four transverse diaphragms. We hypothesized that all fascial restrictions in the study population could be explained by a single underlying factor.
An osteopathic physician (K.W.) recorded somatic dysfunctions of the spine (eg, higher shoulder, sacrum, gait) on the new assessment form (Appendix), noting fascial and spinal restriction on a numbered four-point scale, with severity ratings ranging from 0, absent, to 3, severe. Functionally, the spine is divided into three regions: cervical, thoracic, and lumbar. These regions are further divided into seven subregions based on the anatomic changes in the spinal curves. The cervical spine has two subregions, an upper (C1-C2) and a lower (C3-C7); thoracic, an upper (T1-T4), middle (T5-T8), and lower (T9-T12); and lumbar, an upper (L1-L3) and a lower (L4-L5). As with fascial restrictions, we hypothesized that spinal restrictions were related and could be explained by a single underlying factor of spinal motion restriction. In addition to the structural assessment, the examiner (K.W.) used a visual analog scale to rate the severity of each child's muscle spasticity compared to all other patients with this diagnosis encountered during approximately 15 years of clinical experience.
We hypothesized that the latent factors of restriction in the fascial planes and in the spine would covary and that each of these would be correlated with the examiner-assigned rating for muscle spasticity.
We entered data into a Microsoft Access 2000 for Windows database (Microsoft Corporation, Redmond, Wash). Statistical analysis was conducted using SPSS statistical software (version 11.0 for Windows; SPSS Inc, Chicago, Ill)
20 and EQS structural equation modeling software (version 6.1 for Windows; Multivariate Software, Inc, Encino, Calif).
21 The few instances of missing data were imputed using multiple regression techniques.
We used confirmatory factor analysis to examine the underlying structure of the osteopathic indicators.
21 In this type of analysis, researchers specify the factors a priori (ie, osteopathic findings and the degree of spasticity) and then confirm whether a relationship between the theorized factors does in fact exist. The fit of the theorized model is compared to actual data.
Robust maximum likelihood estimation was used. Model fit statistics reported here are the Satorra-Bentler scaled χ2 statistic, comparative fit index (CFI), and root mean square error of approximation (RMSEA). The χ2 test assesses the size of the discrepancy between the original sample covariance and the proposed model. Statistically significant results from χ2 analysis indicate that the proposed model is significantly different from the data. That is, the a priori model does not explain the actual data well. The CFI compares the proposed model to a null model with values that range from 0 to 1. A CFI higher than 0.95 indicates a good fit between the proposed model and the data. A RMSEA is a measure of misfit by degrees of freedom. Values below 0.10 are considered acceptable; below 0.05, good. Values that are over 0.10 are considered a poor fit. Fit indices are complimentary, the χ2 is sensitive to sample size, the CFI is not as sensitive to sample size, and the RMSEA makes adjustments for the number of model parameters. With a sample size of 57 and an alpha (α) level of .05 powered at 80%, a model with 10 path coefficients had sufficient power to detect a large effect size (0.35).