The role of ovarian steroid hormones in mood
Highlights
► Estrogens and progestagens do not contribute significantly to composite moods of healthy women. ► Changes in daily mood do not map on to menstrual cycle phases as determined by hormone levels. ► Stress and physical health contribute significantly to composite positive and negative moods of healthy women. ► Psychosocial factors may need to be controlled for in any study of mood.
Introduction
Ovarian steroids, with the menstrual cycle as their proxy, are often seen as a determinant of mood in women. The late luteal or premenstrual phase with its drop in estrogens and progestagens is perceived as a time of increased irritability and negativity (Richardson, 1995). However, studies examining a possible relationship between the menstrual cycle or hormone levels and mood in non-help-seeking women have yielded inconsistent findings. While some studies have shown increased negative mood during the premenstrual phase (Freeman et al., 1996, Van Goozen et al., 1997) or decreased positive mood in the premenstrual phase (Cohen et al., 1987) others have shown a cyclical pattern of negative mood that is not confined specifically to the premenstrual period (Hardie, 1997, Sveindottir and Backstrom, 2000). Other studies find no menstrual phase-related changes in mood at all (Almagor and Ben-Porath, 1991, Charette et al., 1990).
Differences in methods of assessing both mood as well as the menstrual cycle phase may account for divergent results. Discrepancies in premenstrual-linked mood changes have been shown to depend on the scale used (Natale and Albertazzi, 2006). Studies that use the menstrual cycle phase as a proxy for directly measured ovarian hormone levels may fail to capture variability within predefined phases (Bellem et al., 2011, Hampson and Young, 2008). Such studies are also liable to unknowingly include anovulatory cycles, resulting in menstrual cycle phases that are not representative of actual hormone levels.
Few studies, in fact, examine directly the relationship between daily hormones and daily mood. Often hormones or mood are only measured on a subset of days during the menstrual cycle (Abplanalp et al., 1979, Fox et al., 2008, Laessle et al., 1990, Rapkin et al., 2011, Stoddard et al., 2007). Composite measures of hormones, such as averages, may be used instead of directly assessing the relationship between daily hormone levels and daily mood (Stoddard et al., 2007). The few studies directly comparing daily measurements of hormones and mood in non-help-seeking women for the duration of at least once complete menstrual cycle have in fact not found a correlation between either estradiol or progesterone and mood (Redei and Freeman, 1995).
Failure to recognize a time lag between hormone change and resulting mood changes may lead to divergent results. Genomic steroid actions are, by definition, slow and prolonged (McEwen and Alves, 1999), suggesting that hormones measured the same day as mood may not reflect their impact on mood until a few days later. Meaden et al. (2005) found a worsening mood during the first 1 to 2 days of menses, rather than in the premenstrual period and explained these results by hypothesizing a lag between hormonal changes and resulting mood symptomology.
Given the divergent results of methodologies and outcomes generally pointing away from a direct influence of ovarian steroids on mood, we wondered whether psychosocial factors might play a stronger role in determining mood. In order to address this question we blinded participants to the object of the study and assessed the relationship between daily levels of estrogen and progesterone metabolites with 12 individual daily mood items – some positive as well as negative – as well as the composites of positive and negative mood items, in a randomly recruited community sample of non-help-seeking women. To better understand the role of psychosocial factors in women's mood we included three other measures in our models: Social Support, Stress and Physical Health. These have been shown to have substantial influence on mental health (Helliwell and Putnam, 2004, McEwen, 2001, Romans et al., 2009) and were used as comparison variables against which to assess the strength of the association between ovarian hormones and daily mood. We included additional models to explore whether hormone levels from 1 to 5 days prior might also influence daily mood. Finally, we explored whether any hormone–mood relationships could be consistently mapped onto menstrual cycle phases, as determined by direct measurement of estrogen and progesterone metabolites.
Section snippets
Mood in Daily Life study
The present study is a sub-study of a larger project, Mood in Daily Life (MiDL) which aimed to ascertain the relationship of daily moods to the menstrual cycle over 24-weeks in a non-help-seeking, community sample of women aged 18–40, living in the Greater Toronto Area in Ontario, Canada. Participants were recruited using a random digit dialing service that called each phone number up to seven times. Interested respondents were asked to attend an initial face-to-face interview, in which
Participants
Twenty-one women participated in the mood and ovarian hormone study. Two participants taking oral contraceptives (assessed via self-report in the initial interview and confirmed via hormone assays) were excluded from all analyses. Thus, 19 participants were included in the LMM analyses assessing the relationship between daily hormones and daily mood. These women ranged in age from 18 to 41 (M ± SD = 33.68 ± 5.50).
From the 19 in whom we assessed the relationship between daily hormones and mood, 5
Discussion
We studied the effects of both ovarian hormone systems on daily mood over one and a half menstrual cycles in a random community sample of non-help-seeking women. We measured estrogen and progesterone metabolite levels daily and examined their contribution as well as that of perceived stress and physical health acquired daily over the same cycle, and social support acquired weekly, to both positive and negative mood items and their composite measures. Importantly, participants were blinded to
Conclusion
In sum, we found that ovarian hormones, as assessed by urinary metabolites E1G and PdG, contributed negligibly to positive and negative mood. In addition, we found no evidence of a relationship between menstrual cycle phase and mood. Rather, perceived stress and physical health were the strongest contributors to daily mood. Taken together, our findings suggest that natural fluctuations of ovarian hormones do not contribute significantly to variations in the daily moods of healthy women and that
Acknowledgments
The authors would like to thank the following individuals for their help in designing and implementing the Mood in Daily Life Study: Nili Benazon, Rose Clarkson, David Kreindler, Sheila Laredo, Anthony Levitt, Katherine Morgan, Michele Petrovic, Donna Stewart and Brenda Toner. We also thank Astrid Bellem for her expert assistance on the hormone enzyme assays and help on the menstrual cycle phase divisions, Drs. Peter Bevan and Elizabeth Page-Gould for their assistance in data analysis, and two
References (41)
- et al.
Mood changes during the menstrual cycle and their relation to the use of oral contraceptive
J. Psychosom. Res.
(1991) - et al.
Measuring estrogens and progestegens in humans: an overview of methods
Gend. Med.
(2011) - et al.
Food cravings, mood, and the menstrual cycle
Horm. Behav.
(1987) - et al.
The presence of both an energy deficiency and estrogen deficiency exacerbate alterations of bone metabolism in exercising women
Bone
(2008) - et al.
Reliability and validity of a daily diary for premenstrual syndrome
Psychiatry Res.
(1996) - et al.
Portable mood mapping: the validity and reliability of analog scale displays for mood assessment via hand-held computer
Psychiatry Res.
(2003) - et al.
Mood changes and physical complaints during the normal menstrual cycle in healthy young women
Psychoneuroendocrinology
(1990) - et al.
Timing and severity of symptoms associated with the menstrual cycle in a community-based sample in the Midwestern United States
Psychiatry Res.
(2005) - et al.
Distribution of positive moods in women with the premenstrual syndrome and in normal women
J. Psychosom. Res.
(1995) - et al.
Neuroimaging evidence of cerebellar involvement in premenstrual dysphoric disorder
Biol. Psychiatry
(2011)