This study is the first to evaluate the links between PA and menopausal HFs using both objective and subjective measures. A key finding is that an acute bout of moderate intensity aerobic exercise decreases both objective and subjective HFs. Although overall PA was not associated with HF frequency in longitudinal analyses over a 15 day period, lower fit women reported more vasomotor symptoms on days when they engaged in more moderate intensity PA than usual. These findings extend our understanding of the relationship between PA and vasomotor symptoms, given the absence of sufficiently powered randomized controlled trials11
and the lack of studies corroborating self-report assessments with objective measures of both PA and vasomotor symptoms.
Our study demonstrated that an acute bout of exercise reduces both objectively and subjectively measured symptoms. This finding is inconsistent with the suppositions of the thermoregulatory model of HFs which suggests that more HFs should follow periods of activity that increase body core temperature such as moderate intensity exercise. The only other study that directly evaluated acute effects of exercise on objectively measured HFs was conducted by Freedman & Krell13
and noted a triggering of HFs following exercise in a laboratory setting. The most likely explanation for the differing results involves the overall level of HFs in our sample which was rather low as compared to the Freedman & Krell study. Interestingly, the calculated effect sizes in the present study indicate it was the women in the two lowest symptom burden categories that experienced symptom reductions, with women in the highest category showing little change (ds = −.25, −.47, and −.04, for low, moderate, and high symptom burden, respectively). Notably, the average number of HFs reported on the PDA across 15 days in the highest symptom burden category was 2.4 HFs per day, which is about half the symptoms used as a criterion for entry into the Freedman & Krell study (≥5 HFs per day) and is also below the FDA definition of moderate to severe symptoms (7–8 HFs per day).36
Additionally, the Freedman & Krell study differed from our study in target population (postmenopausal women versus peri and postmenopausal in our study), environment (HF monitoring in controlled laboratory environment versus in real-life setting in our study), mode and duration of exercise (cycle ergometer at 40% predicted VO2
until either a HF occurred, 60 minutes passed, or participants requested termination; versus a standard 30 minute exercise session on a treadmill in our study). The intensity of exercise was nonetheless comparable between the two studies when expressed as percent predicted
. Future studies are needed to evaluate potential increases in symptoms following exercise in women who are highly symptomatic. More symptomatic women may be also disproportionately represented by overweight (as opposed to normal weight) and late perimenopausal women, the two groups of women that tended to exhibit smaller reductions in subjective symptoms in our study when controlling for each other and the effects of other characteristics (i.e., fitness and symptom burden).
Our study also examined the relationship between daily (habitual) PA and self-reported HFs. At the between-person level, there was no association between any of the PA indicators and HFs. However, in within-person analyses time spent in moderate intensity activity was positively associated with subjective HFs among lower fit women. The effect of daily PA and HFs did not significantly vary by BMI or menopausal status. The involvement of cardiorespiratory fitness in the mechanisms of the association between PA and HFs is plausible given the numerous adaptations endurance trained persons exhibit in thermoregulatory control as compared to untrained persons.37
In addition to attenuated blood pressure and heart rate responses to exercise, aerobically-trained individuals exhibit blunted sympathetic vasomotor activation as compared to untrained individuals.38
Sympathetic activation is higher in symptomatic as compared to asymptomatic menopausal women; drugs that increase sympathetic activation provoke HFs and drugs that decrease this activation ameliorate them. It is believed that an increase in sympathetic activation narrows the thermoneutral zone, which increases the likelihood of HFs, whereas decreased sympathetic activation widens it.15
Additionally, decreases in cardiac vagal control were observed during HFs 39
and aerobically fit/endurance trained individuals have higher heart rate variability (an indirect measure of increased cardiac vagal control) than less fit/sedentary individuals.40
Given these effects involving both sympathetic and parasympathetic control, cardiorespiratory fitness may protect women from experiencing HFs in response to chronic PA.
Our analysis regarding the effects of daily PA on HFs evaluated only self-reported symptoms. Unfortunately, most devices that measure HFs objectively do so for short periods of time (24–72 hours), require that sensors be replaced every 24 hours (e.g., Biolog), or their validity has not been widely established (e.g., FlashMarkPro41
). It should also be noted that any objective monitoring technology may be limited for evaluation of vasomotor symptoms during PA. As most rely on measurements of sweating responses or humidity detection, they are by design unsuitable for use during exercise or periods of PA that result in profuse sweating. Self-report measures thus remain the best option for PA researchers but should be conducted in real time (e.g., using electronic diaries) to minimize recall bias. Self-reported HFs have the additional advantage of face validity; they reflect symptoms that matter to women. In fact, just as in other studies,42
objective and subjective HFs were highly discordant in our study, with significant number of objective HFs not reported by women or reported by women but not objectively detected. Potential explanations for this discordance may include individual differences in somatosensory sensitivity or affect 43
, but more studies are needed to help disentangle the reasons behind under-reporting as well as to improve sensitivity of objective measures of HFs.44
Our sample was well-characterized in terms of health, PA, vasomotor symptoms, and menopausal status, however the women were mostly non-Latina White, educated, and of above average socioeconomic status. Future studies should incorporate more ethnically diverse samples to allow for broader generalizability, especially given ethnic and racial differences in symptom experiences and reporting 45
and in rates of PA.46
Although this is the first study to evaluate changes in objectively and subjectively measured symptoms as a function of acute exercise, our baseline monitoring session did not immediately precede our acute exercise bout. Given the high variability in symptoms, it would be useful to monitor HFs objectively for longer than 24 hours and as close to an acute exercise bout as possible. It should also be noted that the within-person association between moderate intensity activity and HFs may be dependent on the operational definition of moderate intensity activity used. We utilized previously established criteria,47
however other criteria can be used to classify intensity levels. For example, the criteria established by Matthews48
specify moderate intensity as activity level involving accelerometer counts between (760–5724 counts per minute), a wider range than proposed by Freedson (1953–5724 counts per minute considered as moderate). The within-person association was no longer significant when moderate intensity PA was defined using the Matthews criteria. Thus, it is important to interpret our findings as specifically reflecting intensity corresponding to 1953–5724 accelerometer counts per minute.