Background and Purpose. There have been no previous studies that quant
itatively assessed the load on the spine during extensor exercises. Th
e purpose of our study was to investigate the loading of the lumbar sp
ine and trunk muscle activity levels while subjects performed typical
trunk extensor exercises. Subjects. Thirteen male volunteers (mean age
= 21.0 years, SD=1.0, range=l9-23; mean height= 176.0 cm, SD=6.2, rang
e=165-188; mean mass=77.0 kg, SD=7.0, range=63-89) participated. Metho
ds. The subjects performed four different back exercises. Electromyogr
aphic (EMG) activity was recorded from 14 trunk muscles The postures t
hat corresponded external moment were identified and quantified using
rigid body modeling combined with an EMG-driven model to determine joi
nt loading at the L4-5 joint. The exercises were then evaluated based
on the lumbar spine loading-and peak muscle activity levels. A referen
ce task of lifting 10 kg from midthigh was included for comparison. Re
sults. The exercises involving active trunk extension produced the cs
and muscle activity levels. Exercises involving leg extension with the
spine held isometrically demonstrated asymmetricaI activity of tile t
runk muscles, thereby reducing-loads on the spine. Conclusion and Disc
ussion, The back extensor exercises examined provided a wide range of
joint loading and muscle activity levels. Single-leg extension tasks a
ppear to constitute a low-risk exercise for initial extensor strengthe
ning, given the low spine load and mild extensor muscle challenge. Whe
n combined with contralateral arm extensions, the challenge and demand
of the exercise were increased. The compressive loading and extensor
muscle activity levels were highest for the trunk extension exercises.