Effect of age and height on trunk sway during stance and gait

The aim of the studies reported here was to quantify changes in balance control for stance and gait tasks with age and to pinpoint possible advantages and difficulties in using these tasks and measures derived from them to identify pathological balance control in patients. Some 470 normal subjects in the age range 6 to 82 were examined for a battery of 14 stance and gait tasks. During the tasks, angular velocity transducers mounted at lumbar 1-3 measured pitch and roll angular velocities of the body. A combination of outcome measures from several tasks was used to create an overall balance control index. Three types of sensory analyses on pitch angle and velocity amplitudes for stance trials were used to quantify possible changes in the contributions of visual, somatosensory and vestibular inputs to balance control with age for 2-legged stance tasks. Correlation analysis on task variables was used to determine the relationship of subjects' age and height on outcome measures. Outcome measures showed a characteristic "L" or "U" shaped profile with a rapid decrease in values between 7 and 25 years of age, a plateau until 55 then a gradual increase with age after 55 years of age for most stance and gait tasks. The sensory analysis technique using differences between stance tests indicated that visual contributions to balance control continuously increased with age between the ages of 15 and 80, and vestibular and lower leg somatosensory contributions remain relatively constant with age. Sensory analysis calculated as commonly-used quotients of outcome measures revealed large variance across all ages, asymmetric distributions, and no clear trends in sensory contributions to stance with age. A third technique based on a discriminant function analysis using measures from model patient populations indicated that proprioceptive but not vestibular contributions first increased with age and then decreased after 55 years of age. Correlations of outcome measures with age and height indicated that both contributed equally to changes in outcome measures between the ages of 7 and 25, otherwise height had no effect. We conclude that both stance and gait tasks should be selected for identifying changes in balance control from that of healthy persons with a preference for gait tasks as these show less variation with age. Because of the large increases in variance in the elderly and those younger than 20 years, appropriate age-matched reference values should be employed to ascertain if trunk sway is out of normal ranges.