Objective: Our research program focuses on circadian temperature rhythm and adaptation to stress. A feasibility study was done to refine data collection and analysis techniques for a study comparing body temperature (BT) characteristics (1) among phases of the menstrual cycle, and (2) between young, normal cycling women and women transitioning to menopause. After the postovulatory progesterone rise, BT increases in the luteal phase ~ 0.3°C above follicular phase values. In 3 healthy Japanese women followed longitudinally (Nakayama, et al., 1992) and 7 healthy American women (Cagnacci, et al., 1996), amplitude of the circadian temperature rhythm was attenuated significantly in luteal phase compared with follicular phase. Low amplitude suggests an unstable circadian rhythm, indicating lowered resistance to stress and, thus, susceptibility to illness. Circadian temperature rhythm amplitude diminishes with advancing age, but the circadian temperature rhythm in the transition from the reproductive years to menopause has not been characterized.
Design: Cross-sectional (young women) and longitudinal (perimenopausal woman) observational time series design
Sample, Setting, Years: A convenience sample of 3 women, 19-25 years of age, were studied between 1999 and 2001; racial composition was 1 white, 1 black, and 1 Native American. Subjects were free from routine medications except for oral contraceptives (n=2), and one subject was breastfeeding 3 months postpartum. One white perimenopausal woman was studied in years 2000 and 2001 when she was 52 and 53 years old. All measurements were taken between June and August in Houston while subjects were living their normal lives. The institutional review board approved the study.
Concepts/Variables Studied: Stress was conceptualized as: (1) luteal phase of the menstrual cycle, the time when many women experience premenstrual symptoms; and (2) perimenopause, the period of years when many women experience cognitive and physiologic changes. Estrogen affects the organization and expression of circadian rhythms, which help regulate reproductive physiology and behavior. Circadian temperature rhythm is a robust marker of the biological clock’s functional organization. Thus, changes in circadian temperature rhythm are expected to coincide with short-term changes in estrogen production during the monthly menstrual cycle and with long-term changes during perimenopause.
Methods: BT was measured every minute, 24 hrs a day, on 12-40 consecutive days in the young women and on 30-34 consecutive days in the perimenopausal woman. A portable data logger connected to a skin thermistor probe secured to the lower outer quadrant of the abdomen was used to measure temperature. Combined measurement error for the logger and probe was less than 0.15°C. Subjects recorded sleep/wake time, onset and end of menses, meals, medications, alcohol use, exercise, and other activities in a daily log. Onset of menses was retrieved from the log and the highest temperature peak was located between 12 and 16 days before menses onset; the peak was considered ovulation and the time series was subsequently demarcated as follicular, luteal and menstrual phases. When multiple data sets per subject were available, autocorrelation was used to select the data segments with strongest cyclicity. Cosinor analysis was used to determine the period, amplitude, and mesor with an approach that maximizes the cosine amplitude.
Findings: Menstrual phase data were not available for two young women; therefore, the analysis included follicular and luteal phase data. Of the 7 data sets, 5 showed an increase in mesor (0.03° - 0.61°C) in the luteal phase compared with the follicular phase. The postpartum woman had no change in mesor and one data set from the perimenopausal woman showed a 0.13°C decrease. All subjects had circadian temperature rhythm in the follicular and luteal phases. Six data sets had decreased amplitude (0.01° - 0.40°C) in the luteal phase compared with the follicular phase, with the smallest decreases observed in the perimenopausal subject who had a 0.08°C increase in one data set. Amplitude changes in the perimenopausal subject were progressively unstable, with lower values in 2001 compared with 2000.
Conclusions: Although estrogen levels were not measured, circadian temperature rhythm parameters followed theoretical expectations using proximal skin BT measurement and the maximal cosinor approach.
Implications: If the major study findings, when controlling for advancing age, are similar to those of the feasibility study, it will be useful to explore intervention and prevention strategies (e.g., light therapy, rigid sleep/wake routines, exogenous melatonin) to stabilize circadian temperature rhythm. Correlation of stable and unstable rhythms with premenstrual and perimenopausal signs and symptoms may suggest that stabilizing circadian temperature rhythm can improve the well-being of young and middle-aged women.
Funded by PARTNERS, University of Texas-Houston School of Nursing, and Sigma Theta Tau International, Beta Beta at Houston Chapter.
Back to Womens Health Studies: Responses and Risks
Back to The Advancing Nursing Practice Excellence: State of the Science