Interpretation
We observed increased straightening of the coccyx within the pregnant subgroups that had previously delivered solely via CD. Thus, the potential confounding influence of vaginal birth-related injury was absent. This suggests that additional remodeling is achieved in subsequent pregnancies and that these shape differences are likely due to mechanical and hormonal changes of pregnancy alone, and not delivery. This is supported by previous work showing that subsequent pregnancies yielded an increased cellular and hormonal response15. While this result was not significant and may be better delineated with an increased sample size and longitudinal data in a future study, it nevertheless supports the major finding of this study: the coccyx is undergoing significant changes in shape during pregnancy either due to remodeling of the coccyx itself, the tissues attached to it, or both.
Though it has been previously noted that the coccyx moves during vaginal delivery due to direct interaction with the fetal head7,16, this study found that the coccyx also moves in pregnancy by rotating posteriorly about the sacrum presumably in preparation for delivery. It is possible that this straightening reflects relaxation of the levator ani, anococcygeal raphe, and coccygeus (which all insert on this structure); remodeling of the connective tissues supporting the sacrococcygeal joint; or a combination of these events. Straightening of the sacrum-coccyx may be a maternal adaptation to lessen the severity of stretch induced injury that has been demonstrated in prior simulations of vaginal delivery and imaging studies following vaginal delivery17,18. The effects of the growing fetus resulting in increased intraabdominal pressure would likely exhibit a similar effect and could be an alternative contributing mechanism.
Previous studies describe levator ani defects in parous women and there was concern about such injuries impacting the results of this study19–21<sup>21</sup>(21).While some pregnant women in our study may have delivered abdominally after entering the second stage of labor, it is important to note that our pregnant group was vaginally nulliparous and CD in any stage of labor has been found to be protective of levator injury and the development of pelvic floor disorders22,23. Thus, the impact of levator damage was minimized. We did not observe any differences comparing nulliparous and parous women, which would suggest that either no injuries were present in our parous cohort or that the coccyx is not a good indicator of injury.
Previous research by our group evaluating midsagittal pelvic floor shape found that the levator plate was more posterior, or “relaxed”, in pregnant women compared to nulliparous and parous women, coinciding with the posterior motion of the tip of the coccyx resulting in rotation about the sacrum noted in this study24. While it is not clear whether remodeling of the coccyx allows for the noted relaxation or vice versa, the combined implications are that these maternal changes are necessary and more favorable for vaginal delivery. Women that fail to remodel sufficiently may be at greater risk for a complicated delivery and/or maternal injury. Additionally, the failure to recover the pre-pregnancy shape of the coccyx could be an indication of injury either to the coccyx or the muscles that insert on it. While there are currently no clear indications of impaired levator muscle function in the women analyzed in this study, further research may provide indication of impaired levator function based on the shape and/or orientation of the coccyx. Finally, this work indicates that changes in the levator ani/plate may not be solely related to muscle injury. If changes in the coccyx persist long-term, this could alter the length-tension relationships and normal physiologic function of muscles that insert onto it and potentially compromise support to the pelvic organs. While much of these discussion points will require more evidence, this study highlights that the coccyx may be playing a more important role than has been previously considered.
These findings also have implications for computational models of vaginal delivery. We have shown that there is significant remodeling and motion of the coccyx about the sacrum during pregnancy7. If a model of vaginal delivery did not account for these changes (i.e. using nulliparous anatomy) then the geometry may not accurately predict outcomes of vaginal delivery. By not accounting for the remodeling of the coccyx or pelvic floor, that model is likely simulating a pelvis that is not fully prepared for vaginal delivery, meaning the coccyx and levator plate would need to move further posteriorly to reach the same final configuration as a simulated pregnant pelvis.