Biomechanical Evaluation of Head Kinematics during
Infant Shaking vs Pediatric Activities of Daily Living

John  D. Lloyd, Ph.D., M.Erg.S., CPE, CBIS
Board Certified Ergonomist & Certified Brain Injury Specialist

A biomechanical evaluation was performed to quantify kinematic variables associated with various infant shaking techniques, with comparison to a series of pediatric activities of daily living, as well as determination of injury risk as a function of known biomechanical thresholds of injury.

Methodology
Three infant representatives were utilized for these tasks, including the demonstration doll from the National Center for Shaken Baby Syndrome, a CRABI biofidelic mannequin and a seven-month-old infant male. Tasks were performed by nine subjects, ranging in age from 20 years through 77 years and included two females and seven males. Basic anthropometry of the infant representatives are presented below:

Table 1: Basic anthropometry of infant representatives

Shaking events were investigated using both the NCSBS doll and CRABI mannequin. These tasks included: (i) Resuscitative shaking in the flexion-extension plane, (ii) Resuscitative shaking in about yaw axis, (iii) Harmful shaking, and (iv) Gravity assisted shaking

Figure 3: Male subject #1 performing harmful shaking of NCSBS demonstration doll

Pediatric Activities of Daily Living (PADL) were explored primarily using the CRABI biofidelic mannequin. These activities included: pushing infant in various strollers over smooth and uneven surfaces; walking and running on treadmill while holding the infant in a baby carrier; swinging infant in powered cradle; burping – both back slap and up/down shaking techniques; rough play – throw up in the air and catch; bouncing infant on knee; hitching infant up on the hip; consoling behavior; and swinging back and forth. Additionally, the author’s son, Kieran Lloyd, shared his enthusiasm for bouncing in his Fisher Price jumparoo.

Measures of head kinematics were acquired using sensors attached to the head and torso of the infant representative. Raw data, including orientation (yaw, pitch, roll), quaternion, angular velocity and linear acceleration were transmitted wirelessly to a nearby computer.

Using MatLab (The MathWorks. Natick, MA), Fast Fourier Transform (FFT) analysis was performed to identify and isolate true data from environmental noise, which was subsequently removed using a 4th order Butterworth low-pass filter. Angular accelerations were subsequently derived, RMS values calculated and Head Injury Criterion (HIC15) computed.

Summary of Results

The key result from this study is that Head Angular Acceleration is very similar between intended harmful shaking of the CRABI biofidelic mannequin (1068.3 rad/s2) and a 7 month old infant male playing in his jumparoo (954.4 rad/s2). Furthermore, measures of head angular acceleration are substantially below scientifically accepted biomechanical thresholds of injury.

Thus, shaking produces head kinematics that are clearly benign and well tolerated by normal infants, even if repetitive. If intracranial injury (SDH, EDH, SAH, DAI, concussion) is clinically presented, our data would indicate that shaking would not be part of its etiology.

In summary, this biomechanical study presents several important results.

• Firstly, that infant shaking produces head motions that are far below the levels required to cause injury in children
• Secondly, head motion during vigorous shaking is very similar to head motion while a 7-month-old infant plays in his jumparoo and there are no recorded cases of head injury attributed to such playful activity
• Thirdly,playing in the jumparoo causes repetitive head motion of a frequency similar to that experienced during shaking. Thus repetitive head motion does not increase risk of brain injury.
• Finally, that the NCSBS demonstration doll is an unsuitable infant representative.