Depressive episodes in premenstrual syndrome.

Episodic depression, a prominent but poorly defined symptom of premenstrual syndrome, was quantitated in 24-hour cortisol secretory episodes (determined by sampling at 20-minute intervals) as biochemical markers, as well as the Beck Depression Inventory and Profile of Mood States as psychometric measures. Results of 16 patients with premenstrual syndrome were compared with six age-matched women with endogenous depression and 16 control women. On both the Profile of Mood States and Beck Depression Inventory, women with premenstrual syndrome showed a marked worsening of scores (p less than 0.01) during the luteal phase compared with either their own follicular phase scores or the scores of controls in either cycle phase. However, Beck Depression Inventory scores were threefold higher (p less than 0.005) in women with depression than in those with luteal phase premenstrual syndrome (3.37 +/- 3.6 vs. 11.9 +/- 2.5). The Profile of Mood States depression scale was also higher (p less than 0.05) in women with depression than in those with premenstrual syndrome, while scores on other Profile of Mood States scales were similar. The numbers of cortisol secretory pulses identified by the cluster algorithm were similar (5 to 6 per 24 hours) in all groups, and the time of circadian nadirs as determined by cosinor rhythmometry were comparable. While the mean amplitude and duration of the cortisol pulses were also similar in women with premenstrual syndrome and controls, both were significantly higher (p less than 0.05) in women with depression. This resulted in markedly enhanced (p less than 0.005) cortisol secretion during a given secretory episode in women with depression and in higher 24-hour transverse mean cortisol values in women with depression (87.8 +/- 5.8 ng/ml) than in either those with premenstrual syndrome (66.7 +/- 3.3 ng/ml) or controls (58.9 +/- 3.3 ng/ml). These data affirm the clinical impression that depressive episodes occurring selectively in the luteal phase of the cycle in women with premenstrual syndrome are not present in controls and demonstrate, for the first time, that these episodes are distinct from endogenous depression as measured by both cortisol secretory parameters and psychological indices.