I do not think there is much doubt that stress can alter markers of inflammatory activation. In baboons, stress associated with animal transport results in a very high rate of pregnancy loss.
That said, we should be careful in attributing adverse pregnancy outcomes due to preeclampsia in human pregnancy to maternal stress without clear data. The association opens the door for women to blame themselves for having too much stress in their lives and therefore being responsible for their bad outcome. A number of us have tried to find cytokine markers that would serve as markers or predictors of preeclampsia - without much success. There are may be some statistical associations - but the overlap in data point is usually very large.
The abstract is very general in its content. I would certainly want to see more about study design and the robustness of the findings.