Sleep health and race: little-chartered territory

In this issue of SLEEP, analyzing cross-sectional US national health survey data spanning a 14-year period, Wang et al. [1] report on trends in self-reported prevalence (distinguishing by sex and racial grouping) of two common sleep problems: short sleep time (defined as <6 h of sleep) and trouble sleeping. Wang et al. [1] found that while short sleep time was relatively commonly encountered in the US population (ranging anywhere from 13% to 37%, depending on the year of evaluation and how sleep time was asked about), there was no significant change in the population prevalence of short sleep time between 2005 and 2014 (and during those years, a consistent method of sleep time ascertainment was applied). In contrast, a progressive and significant increase in the population prevalence of trouble sleeping was observed between 2005 and 2018 (and trouble sleeping was ascertained by a consistent method throughout those years), initially 23.7% in 2005 and rising to 28.7% by 2018 [1]. Some of Wang et al.’s [1] more intriguing findings though relate to associations between sleep problems and race. Between 2005 and 2018, short sleep time was consistently more frequently observed among non-Whites (and particularly non-Hispanic Blacks) versus Whites, whereas trouble sleeping was uniformly more commonly reported by Whites versus non-Whites [1]. It should be acknowledged though that in a multivariable adjusted analysis (found in the Online Supplement), non-Whites had in fact significantly greater independent odds of experiencing trouble sleeping over Whites [1]. The finding of greater frequency of trouble sleeping among Whites versus non-Whites by Wang et al. [1] provides more insight into a previously recognized serious public health problem with racial dimensions: Whites are more likely to use sleeping pills compared to non-Whites [2–7]. While insomnia clinical practice guidelines recommend psychological and behavioral interventions as first-line management for insomnia [8], health care providers often resort to prescribing sleeping pills (like nonselective benzodiazepine receptor agonists, selective benzodiazepine receptor agonists [Z-drugs], etc.) for insomnia, because of limited success with the aforementioned non-pharmacologic strategies. However, sleeping pills are in fact only marginally effective at improving sleep quality [9, 10], and their use is associated with a broad range of serious adverse events, including abuse [11, 12], withdrawal symptoms [13], impaired cognitive function [9], dementia [14, 15], motor vehicle accidents [16, 17], falls [18, 19], fractures [20], and pneumonia [21, 22]. Therefore, while chronic insomnia can be associated with important consequences on level of function, health, and quality of life, health care providers’ response to the issue (i.e. the prescription of sleeping pills) is likely even more problematic. Multiple US studies have identified that Whites are at significantly greater risk of receiving sleeping pills compared to other racial groups [2–7], even after adjusting for covariates like income and education level, and insurance status [3, 5, 7]. Particularly concerning sedative drug usage patterns, including concurrent receipt of a benzodiazepine and an opioid, and co-administration of a nonselective benzodiazepine and a selective benzodiazepine, also more frequently occur among Whites versus other racial groups [7]. Connected to the topic of sleeping pill use, it has been recognized that the North American “opioid crisis,” which has captured much mass media attention, has been disproportionately driven by Whites [23, 24]. It is noteworthy, and even surprising, to observe that Whites, who are consistently at lower risk for a variety of adverse health outcomes compared to other racial groups [25], to be at greater risk for this sleep medicine-related negative heath metric (i.e. the receipt of sleeping pills). While disparities in health care professional management of insomnia and differences in treatment-seeking behavior among racial groups are plausible explanations for the greater receipt of sleeping pills among Whites, Wang et al.’s [1] study results identify that insomnia itself appears to more commonly and consistently affect Whites than non-Whites, thereby potentially driving the increased sleeping pill usage in the former group. Why insomnia may be more prevalent among Whites is unclear and needs further research. Possible factors include differing perception and articulation of trouble sleeping, differing lifestyle and behaviors that influence sleep, and differing comorbidities, among races. Although trouble sleeping was more commonly reported among Whites across the 14-year study period, curiously, short sleep time was not, and it was instead was more frequently and consistently experienced by non-Whites, and particularly non-Hispanic Blacks [1]. Wang et al. [1] propose several possible reasons for the higher frequency of short sleep time experienced among non-Whites versus Whites, including differences in work patterns, work stressors, and comorbidity profile. The finding of more prevalent short sleep time, but not trouble sleeping (and one might reasonably anticipate that the two sleep problems would overlap), among non-Whites versus Whites is somewhat of a curious discrepancy. One possible explanation for this discrepancy might be that insomnia indeed occurs more frequently among non-Whites than Wang et al. [1] have estimated, but there might be a differing perception and articulation of trouble sleeping (that does not extend to estimation of sleep time) by racial group. Indeed, Wang et al.’s [1] study shows just how influential the wording of questions are for evaluating sleep health. While trouble sleeping was evaluated the same way across the seven NHANES cycles analyzed, the wording of the sleep time question changed between the 2005–2014 and 2015–2018 periods. When sleep time was estimated in the 2015-2018 period by calculating the difference in self-reported bed and rise times (versus in the 2005–2014 period, when it was evaluated by directly asking how much one usually slept on weekdays or workdays), a dramatically ~60% lower population prevalence of short sleep time was found [1]. While Wang et al [1]. provide necessary and informative population-level prevalence trends in short sleep time and trouble sleeping, distinguishing by sex and racial grouping, several limitations need to be taken into consideration when interpreting their results. Short sleep time, and particularly trouble sleeping, were crudely estimated by the NHANES questions. The term “trouble sleeping” is extremely vague, as it can take a variety of specific forms (including difficulty initiating sleep, recurrent nocturnal awakenings of brief or prolonged duration, and undesired early morning awakening), and it can be as a result of various pathologies (e.g. sleep breathing disorder, sleep movement disorder, and so on), and none such details are elucidated by the Wang et al. [1] study. The NHANES questions and categorization of survey participants’ race (i.e. non-Hispanic White vs. non-Hispanic Black vs. Mexican American vs. other Hispanic vs. other race) are also crude and categorization may be particularly challenging for individuals of multi-racial background. In summary, Wang et al. [1]. report intriguing and somewhat challenging to reconcile racial discrepancies in sleep problems: trouble sleeping (but not short sleep time) is more commonly experienced by Whites, and short sleep time (but not trouble sleeping) is more frequently reported by non-Whites. Additional research is needed to confirm these findings and to better appreciate the factors that are contributing to these possible relationships. A clearer understanding of the possible associations between sleep duration, insomnia, and race will facilitate superior management of these aforementioned common sleep-related problems for all people. None declared.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.