Symptom experiences of midlife women: observations from the Seattle Midlife Women's Health Study

Symptoms experienced by perimenopausal women are varied with little agreement about their nature, cause or stability.

To describe the type and stability of symptoms experienced by midlife women.

A community-based sample of 301 women ages 35-55 (Mean 41.1; S.D. = 4.2), had at least one menstrual period in the past year, and took no ovarian hormones. Women completed a daily symptom diary for at least one cycle for 3 consecutive years. Symptoms were rated from zero to four. The 5 premenses days were targeted for consistency and as those most symptomatic. Twenty-eight symptoms commonly reported as perimenopausal were factor analyzed using principal components analysis with varimax rotation. Test-retest reliability and stability estimates were calculated according to the method of Heise (Heise, D. Am Sociol Rev 1969; 34: 93-101) that accounts for expected change over time.

Twenty-five of the 28 symptoms loaded on five factors labeled dysphoric mood, vasomotor, somatic, neuromuscular, and insomnia together accounting for 51.7% of the variance. Test-retest reliability estimates were highest for dysphoric mood (r = 0.78) and somatic (r = 0.70) symptoms. The reliability for the other three clusters ranged from r = 0.65 to r = 0.53). The stability of the clusters across 3 years was high for dysphoric mood, neuromuscular, and insomnia. The vasomotor and somatic clusters had the most change between years 1 and 3.

These results indicate that dysphoric mood is not a part of vasomotor symptoms or insomnia or other somatic symptoms suggesting an origin for vasomotor symptoms apart from the other symptoms. The stability of dysphoric mood across 3 years suggests a chronic situation possibly due to high stress, overwork, or an ongoing emotional illness. The stability of the neuromuscular and insomnia symptoms suggests underlying chronic physical conditions. The reduction in stability of vasomotor symptoms may reflect the changing nature of hormones as women approach menopause. Finally, the low stability of somatic symptoms suggests that they represent acute episodic illnesses. Together the identification of five distinct symptom clusters with varying stability over 3 years suggests that they are due to different underlying mechanisms and are not all attributed to the changing hormone patterns associated with the menopausal transition. Many other events in a midlife woman's life can account for these symptoms including life stress and acute and chronic illnesses.