Symptoms of depression and impaired endothelial function in healthy adolescent women

Two reviewers of this paper noted that the association between depressive symptoms and endothelial dysfunction could be a result of confounding by menstrual status. Unfortunately, we did not have hormonal samples available to determine exactly where in her cycle each participant was at the time of assessment. However, we did collect self-reports of cycle length, timing, and regularity. Using the methods outlined by Holding and Minkoff (1973), we used these data to categorize participants as currently menstruating versus in the follicular or luteal phases of their cycle. Because some of these data were missing, and other participants reported irregular cycles, these analyses could only be done on about half of the sample (n = 52). Nonetheless, among these participants there was no evidence of a relationship between menstrual status and either depressive symptoms (r = .04, P = .79) or endothelial function (r = −.01, P = .95). There also were no significant associations between these variables and cycle regularity (r’s < .12, P’s > .26). This pattern of findings suggests that menstrual status is unlikely to be acting as a confound in our sample. However, we recognize that our approach to these analyses has limitations, and in future research a more rigorous assessment of menstrual status using hormonal data would be valuable.