Other factors
More pertinent might be indications that HPO hormone changes are a dominant factor in some but not in all women. When women in the Tom et al study were divided by their subjective evaluation of how bothersome their sleep disturbances were evident was the differential impact of menopause status versus other factors, that is, the vasomotor, somatic, and psychological symptoms and waking at night to use the toilet. For the group of women who reported being a little bothered by their sleep (moderate), increased odds of sleep difficulties seen by menopause status disappeared when other concurrently measured factors were taken into account, except in women who had undergone hysterectomies. These results support prior observations that sleep difficulties during the menopausal transition in some women are more related to other changes or sensations. However, after the same adjustments were done for the menopause status groups who reported being a lot bothered by their sleep difficulties (severe), the odds of reporting sleep difficulties were not reduced, suggesting that menopause status was dominantly influential. Women reporting more severe sleep bother might be construed as vulnerable to menopausal ovarian hormone shifts impacting sleep, regardless of other contextual factors. Thus, menopause status (hormone fluctuations) may be the unparalleled contributor to perceived sleep quality in some but not all women during the menopausal transition, and maybe not the majority. According to Tom et al, most women (72%) who reported sleep difficulty fell into the moderate category.

The observations by Tom and her team regarding sleep quality subgroups are different from but parallel to observations we have made of two sleep subgroups of midlife women. We found no influence of menopause status on reported or recorded sleep quality. However, using both modalities and comparing women reporting poor sleep (insomnia) with midlife women reporting good sleep, one group had insomnia that was not corroborated on sleep recordings. They had high psychological distress and general somatic symptom scores, although VMS scores were not comparatively augmented. Another group reporting insomnia that was corroborated with recordings had high self-reported VMS scores but not comparatively high psychological distress or somatic symptom scores. From these data, perceived poor sleep in some women seemed to be more a function of high life strain and psychological distress than ovarian hormone fluctuations, and in others, sleep interference seemed much more probably influenced by menopause status, presumably the hormone shifts in menopause.

Investigative frames
Although much evidence is pursued to declare HPO axis hormone fluctuations as the preeminent culprits in creating menopause-related symptom or behavioral manifestations, the Tom et al study exemplifies the value in broadening the frame for identifying factors that would more accurately predict or explain them across individuals. The emphasis on HPO axis function in explaining menopause manifestations perhaps stems from a common frame for biomedical science, held over from the germ theory idea used early in the origins of understanding infectious disease, that a specific type of microorganism causes a specific disease. Using a biomedical frame, it has been typical to design studies of health-related phenomena as if looking for the “cause” (ie, generalizable single or dominant instigator) of the phenomenon. Implicit in this approach is that if we know the cause, the “cure” will become evident. Even in the realm of common infectious diseases, obviating the completeness of the germ theory were observations that some healthy people carry germs (pathogens) but have no symptoms of disease and that some early conditions, for example, scurvy, were not attributable to germs. Similarly, in this day and age, within industrialized societies, microorganisms are rarely detectable in evolving chronic diseases, that is, not in those with detectable pathologly prevalent nor those that fall in the ever-increasing functional disorder category. These comments are made in no way to denigrate the importance or necessity of what is now the classic biomedical frame but to argue that framing of health science beyond this disease science approach is needed.

Increasingly clear in our society is that most diseases/illnesses are chronic and attributable to no one dominant factor but rather a variety of personal style (eg, temperament) and lifestyle behaviors. People at risk for disease/illness often include those prone to be stress hyperaroused or display aggression or depression or those who smoke, abuse alcohol, overeat, underexercise, lack sleep, or engage in risky sexual behaviors. When explaining phenomena such as sleep, which is a function of people in interface with their environments, salient factors to explaining disturbances are likely to include combinations of “person” vulnerability (both body and mind) factors and “environmental” risk (both physical and social) factors. Thus, more attention in human health science might be paid using a human health ecology frame and going beyond a “one size fits all” approach. The search is for sets of contributing factors rather than any dominant instigator. Indeed, as seen in the Tom et al article, after adjusting for age, a set of ecological factors, including somatic vasomotor and psychological symptoms and waking to use the toilet at night, among others, impacted perceived sleep quality. Moreover, multiple factors in person and environmental realms are likely to be differentially influential in inducing or worsening symptoms or behaviors in people, thereby defining subsets of people by factor configurations, also revealed in the Tom et al study according to the two sleep groups. One can see the challenges and complexities inherent in this approach as it necessitates multivariate data collection and analyses that will reveal multiple clusters of predictive or explanatory factors, thus defining subgroups of people requiring multimodal treatments/interventions tailored to particular configurations of contributing factors.

Sleep problems as newly experienced or exacerbated
With regard to the initial question of whether sleep issues are newly experienced or exacerbated in menopause, Tom et al neglected to comment further in the “Results” or “Discussion” sections, and from the explicated data, few insights emerged. Indicating that many women did not declare preexisting sleep problems, Tom et al showed in their Table 1 that about 17% of the women at age 43 years reported trouble with sleep once or twice a week or 3 to 10 times per month for a minimum of 4 months duration. By age 48 years, nearly 38% of the women were reported to be a little (moderate) or a lot (severe) bothered by trouble sleeping, and this increased by 10% more to 48% by age 54 years. Unclear is whether the 38% at age 48 years includes the 17% of women with longstanding sleep problems and what proportion of them were pre or already in transition to menopause (thought for most to begin by the mid-40s), and chances are good that for many, life factors were impacting their perceptions. A difficulty in interpreting or assessing these data is that the sleep history and the repeated, current sleep quality items are not comparable. The former requires a response according to frequency and the latter requires a subjective evaluative/interpretive response, if poor sleep is perceived. In US samples, insomnia prevalence is estimated to be about 30% at any one time, and about 10% are classed as having enduring insomnia. Thus, the propensity to chronic insomnia is likely to be a factor contributing to menopause sleep manifestations in only a minority of women.