Fig 1.
Upper (UC), lower (LC) and thorax segment (Thx) angles from the 3-segment model (upper panel), and head center of mass migration (HCM, lower panel) for one trial of global cervical flexion (left column), global cervical extension (middle column) and head protraction (right column) for a representative subject in the neck pain sample.
Table 1.
The between- and within-subject variance components of ROM and postural variables calculated from an unpublished test-retest reliability study (N = 13).
The required size of each group was estimated for 20% detectable difference between groups for the HCM_global and HCM_prot.
Table 2.
Descriptive statistics for the control and neck pain groups (mean and SD except for duration that is median and inter quartile range).
The NDI is normalised to the range of 0 to 100.
Table 3.
Descriptive statistics for all kinematic variables separately for the control and neck pain groups (Mean (SD)).
All variables have unit degrees except for the HCM variables that have unit cm.
Fig 2.
Forest plot of parameter estimates of group differences between neck pain participants and controls along with 95% confidence intervals for adjusted (ANCOVA) and unadjusted (ANOVA) analyses of cervical range of motion (ROM).
The covariate was self-rated physical activity. Positive values favour increased ROM in the control group compared to the neck pain participants.
Fig 3.
Scatterplot of the upper cervical angle in the natural head posture (UC_NHP) versus upper cervical ROM in extension (UC_ext) in A, and the lower cervical angle in the natural head posture (LC_NHP) versus lower cervical ROM in flexion (LC_flex) in B. Filled circles represent the neck pain subjects (n = 120) and open circles represent controls (n = 40).
Fig 4.
Parameter estimates of group differences in degrees along with 95% confidence intervals separately for upper cervical extension (UC_ext) and lower cervical flexion (LC_flex) for the models with (ANCOVA) and without (ANOVA) initial posture as covariate.
For UC_ext, a more negative value represents greater impairment in the NP group compared to CON. For LC_flex, a more positive value represents a greater impairment in the NP group compared to CON.
Fig 5.
Interaction plot group × test for HCM migration normalised to the averages for the CON group separately for each test, with 95% confidence intervals.