In this study, the IH of transverse pelvic and thoracic rotations, the effect of walking velocity on phase relationship and coupling strength of between these rotations in healthy nulliparous and healthy pregnant women as well as their group differences were compared to assess the influence of pregnancy on axial rotations of pelvis and thorax and their coordination during walking at various velocities. To our surprise, there is no significant difference were found between the two groups in motion pattern of axial rotations of pelvis and thorax and the gait coordination.
There are some literatures on gait and posture analysis in healthy pregnant women. Some studies have showed a forward displacement of the center of gravity[19-20], and an increase of the lumbar lordosis, accompanied by a forward tilt of the pelvis, while other studies, however, found that the center of gravity to be displaced backwards[21-22], or, sometimes, found no significant displacement[23]. Moreover, Snijders et al[24] reported that body length often increases in pregnancy through a reduction of the lumbar lordosis, while others found no significant change in lumbar lordosis[25], or an actual increase[26]. Alvarez et al[27] showed no pregnancy-related change in the base of gait, while Bird, Menz and Hyde[28] reported it to be increased. These latter authors found no systematic increase in the foot progression angle. On the other hand, Foti, Davids and Bagley[16] reported no change in pelvic obliquity or in the horizontal rotations of the pelvis. So, the most conspicuous finding of this literature is variability in postural adaptations rather than any systematic effect. Pregnant women adapt to changes of pregnancy[16].
It has been reported that on average a pregnant woman gain about 12 kilograms in weight[27]. It has been found the increase of the load on the trunk will influence the coordination between transverse pelvic and thoracic rotations during gait in healthy subjects[29]. Though in pregnancy, not only the mechanical load on the trunk has increase, but also significant changes on the musculoskeletal system may exist. Changes on gait coordination may be expected in pregnant women.
Walking especially at high velocity is a balancing movement demanding fast and accurate coordination of muscles in thorax, lumbar and pelvis, hips, legs and feet. Our study once again showed that the synchronized movements of the spine and pelvis within bipedal locomotion are attained in healthy pregnant women. Considering the physical and physiological changes in pregnancy, it could not reflect in the gait coordination. Our findings support our hypothesis that the healthy pregnant women adapt the changes of pregnancy in terms of gait coordination. This kind of adaptation does not necessarily imply that they are able do this same task by using the same amount of energy. In fact, it has been speculated that pregnant women need more energy to perform the same task compared with controls[15]. Again, here we could not find systematic effect of pregnancy on the coordination between transverse pelvic and thoracic rotations during walking. Chapman and Kurokawa hypothesized that the relations between upper and lower trunk rotation are a “damping phenomenon (which) may contribute to the efficiency of locomotion but are not yet fully understood”[30]. It is almost a tautology of human locomotion research that the motion of the various body segments “cooperate” to facilitate specific gait, minimize forces and or energy expenditure, or ensure adequate stability, or some combination of these[31]. The healthy pregnant women may cost more energy for normal and fast walking although they use the same coordination pattern during gait and they are able walk as fast as the nulliparous healthy women.
Previous work showed that subjects naturally adopt a walking velocity which optimise energy cost of locomotion and stability of stride[32]. Humans seem to have strong tendency to walk in ways that minimise metabolic energy costs. Healthy subjects prefer to walk at speeds, close to the speed (at about 1.4 m/s) at which the energy cost per unit distance is least[33]. In this study, the majority of the pregnant women were able to reach the 6.2 km/h like the control subjects. However, the comfortable velocity in pregnant women was significantly reduced compared with the controls. We speculate that the pregnant women adopt the reduction of comfortable walking velocity in order to economise the energy cost.

Conclusion
The comfortable walking velocity was significantly reduced in healthy pregnant women compared with the healthy controls. There was no significant difference with respect to Tempa Scale, IH of transverse pelvic and thoracic rotations, phase relation and coupling strength between these rotations between the pregnant women and controls. No effect of pregnancy on gait coordination between transverse pelvic and thoracic rotations was found. The pregnant women were able to adapt to the changes by the pregnancy in terms of gait coordination. The pregnant women adopt the reduction of comfortable walking velocity in order to economise the energy cost.

Acknowledgement
This paper is dedicated to the memory of Professor Paul I.J.M. Wuisman. Kimi Uegaki contributed to the writing of an early version of the manuscript.