Is physical activity beneficial for hot flashes?

Image by mikebaird via Flickr

Can physical activity help resolve menopausal hot flashes? In this article, we summarize the reports regarding physical activity and hot flash/night sweats otherwise known as VMS(vasomotor symptoms) and discuss biological mechanisms by which physical activity might exert a beneficial effect. The potentially mediating role of mood, quality of sleep, and/or body weight on the relation between physical activity and VMS will be reviewed, as will the clinical and public health implications of our current state of knowledge.

Review of the literature

Approximately 27 published articles have explored the relation between physical activity and hot flashes (a complete reference list is available from the author). Generally, these studies have evaluated Caucasian populations in the United States, Australia, and Sweden; a few have included African Americans and other racial/ethnic groups.^1-3

Most studies feature observational, cross-sectional designs: one had a case-control design,⁴ and one cross-sectional study assessed physical activity prior to onset of VMS.⁶ Two observational studies followed cohorts prospectively,^1,2 and 6 were randomized controlled trials. Assessment of physical activity ranged from response to a single global question to detailed recalls of activity duration, frequency, and mode.

Exercise interventions in the randomized clinical trials usually featured moderate-intensity walking programs, 30 minutes a day, 3 to 5 days a week for 12 to 16 weeks. One intervention specifically evaluated increased intensity of exercise over time.⁶

Symptom assessment also varied: some studies considered frequency, severity, and/or bother as separate domains; others used a single measure or symptom frequency. No study has objectively measured VMS.

Many observational studies had null findings, but 2 reported significantly increased risk of hot flashes in active women.^5,7 One showed increased risk only in women who were highly active at a younger age.⁵ Other studies have reported protective associations: In one study, the prevalence of moderate to severe hot flashes in women in an exercise program was reported to be 21.5% compared with 43.8% in nonparticipants.⁸ A more recent study⁹ noted that highly active postmenopausal women had a lower prevalence of hot flushes compared with those who had little or no exercise (5% vs 14%-16%; P<.05); however, 35% of the sample used hormone therapy (HT). HT users were more likely to be active than nonusers.

Perhaps the strongest observational data indicating a protective association of exercise on VMS comes from the longitudinal Melbourne Women’s Midlife Health Project, which followed 438 women for 8 years.² At baseline, physical activity was not associated with VMS in this cohort¹⁰; however, women who at study initiation reported exercising every day were 49% less likely to report bothersome hot flashes during follow-up (odds ratio [OR] = 0.51; 95% confidence interval [CI] = 0.27-0.96). Over follow-up, decreases in exercise level were associated with increased VMS.

The results from randomized trials are inconsistent. Two trials (only one was designed to test a specific hypothesis about VMS) reported no effect of exercise on VMS; one reported a significant increase in hot flash severity in exercisers vs controls. In contrast, 2 small, short-term trials reported statistically significant reductions in frequency and severity of VMS. A 4-month intervention enrolling 164 previously sedentary women randomized either to a walking group, yoga, or a control group⁶ showed decreased VMS in both arms relative to the control group; however, the differences were not statistically significant. Change in VMS appeared to be mediated by increases in physical fitness: participants who had the most pronounced improvement in fitness also had the most significant decrease in symptoms.

In summary, although the evidence for a protective effect of exercise on VMS is minimal, the literature is limited. Most studies had insufficient power to detect any potential effect.

Potential biological mechanisms of VMS

Although the etiology of the hot flush is not fully understood, experiments suggest a variety of potential mediators, many of which are associated with exercise.

Studies suggest that an increase in the body’s core temperature precedes the onset of a hot flash.¹¹ Potentially, a narrowing of the thermoregulatory zone triggers the event. This may result from a decrease in the threshold for sweating and/or an increase in the threshold for shivering.¹²

Stress is a precipitating factor for hot flashes¹³; therefore, neuroendocrine substances—which play a role in the stress response and affect thermoregulation at the level of the hypothalamus—may be implicated in their etiology.

Prior to hot flush onset, increased levels of brain norepinephrine (NE) are observed and are further elevated during episodes,¹¹ suggesting that hot flushes may result from imbalances in the autonomic system. The “stress-buffering” role of the parasympathetic nervous system may be inadequate to counter the increased activation of the sympathetic nervous system.^14,15 Heart rate variability—an indicator of reduced vagal control—appears to be reduced during the hot flash itself,¹⁶ providing some empirical support for this hypothesis.

Increased cortisol levels are observed during the later stages of the menopausal transition. Higher cortisol levels are associated with higher epinephrine and NE levels as well as with more severe VMS^17,18 and may also indicate an imbalance in the autonomic nervous system.

Animal studies suggest that b-endorphins may play a role in the pathogenesis of the hot flush. Administration of naloxone, an opiate antagonist, in morphine-dependent rats causes symptoms similar to those of the hot flash, as well as luteinizing hormone (LH) surge.¹⁹

However, in postmenopausal women, the infusion of naloxone has not consistently reduced the frequency of hot flushes or LH pulses.²⁰ Studies of plasma b-endorphin levels prior to hot flushes have also been contradictory,^21,22 although plasma levels may not reflect the endorphin levels in the brain....read more