Brigham Research Institute Poster Session Site logo-1
Close this search box.

Jessica Harder, MD






Brigham and Women's Hospital




Jessica A. Harder, MD, Raina N. Fichorova, MD, PhD, Akanksha Srivastava, BA, Aleta Wiley, MPH, Katherine E. Burdick, PhD, Joseph J. Locascio, PhD, Hadine Joffe, MD MS

Principal Investigator

Hadine Joffe


BDNF is Inversely Associated with Mood in Perimenopausal Depression

This research begins to characterize the distinct biomarker profiles associated with mood disorders sensitive to gonadal steroid hormone changes. This work intrinsically addresses key issues in women’s health by focusing on the impact of these hormonal shifts and distinguishing clinically similar presentations by underlying gender-based biological contributors.


Previous work implicates high pro-inflammatory biomarkers in mood disturbance and low brain-derived neurotrophic factor (BDNF) levels in major depression. However, in hormonally-sensitive premenstrual dysphoric disorder (PMDD), BDNF levels are higher when mood is worse. Perimenopausal depression has not been studied to date. We evaluated whether BDNF and inflammatory cytokines predict mood symptoms across the menstrual cycle in hormonally-sensitive perimenopausal depression symptoms.



Data from 49 time points derived from mid-to-late follicular phase [M/L-FP] and peri-menstrual assessments of 14 perimenopausal women ages 38-52 with ovulatory menstrual cycles 24-35 days long across 1-2 cycles for mood symptoms, BDNF levels, cytokines, gonadal steroids. Depression was assessed with Montgomery-Åsberg Depression Rating Scale (MADRS), Beck Depression Inventory (BDI); irritability with Kellner Symptom Questionnaire Anger-Hostility subscale (SQ); overall psychological distress with Profile of Mood States (POMS). Mixed models were run on dependent measures of MADRS (primary endpoint) and other mood outcomes (BDI, POMS, SQ) with independent variables of interest (each biomarker, cycle phase), controlling for cycle number and participant.



After FDR adjustment, BDNF levels showed consistent significant positive relationships to MADRS (beta=0.00053; p=0.0028), POMS (beta=0.00153; p=0.0394), SQ (beta=0.00053; p=0.0067), and BDI (beta=0.00039; p=0.0231). Cycle phase did not affect this relationship. No other biomarker consistently predicted affective symptom severity.


In women with perimenopausal depression symptoms, BDNF is elevated in association with more severe mood symptomatology, resembling the pattern in hormonally-sensitive PMDD and suggesting a hormonally-sensitive mood disorder biomarker profile distinct from that of major depression.

Research Context