The role of obstructive sleep apnea/hypopnea for leukoaraiosis and its cognitive consequences: a discussion still open!

The neurological and neurovascular impact of obstructive sleep apnea/hypopnea (OSA) is complex and still poorly understood. While OSA appears to be associated with an increased risk of a first stroke and is associated with an enhanced risk of recurrence of this same type of cardiovascular accident [1, 2], the brain microvasculature consequences of this syndrome are less well known.

The results published in the issue of this journal [3] lead to a complex discussion of the (causal) link between OSA and subsequent dementia of vascular origin. In a highly selected population, the authors report the absence of any association (according to cross-sectional and longitudinal analyses) between lesions indicative of cerebral/ophthalmic microvascular fragility and their progression over 3 years of follow-up and also the presence of OSA syndrome. Another determinant information is the absence of such an association irrespective of OSA severity according to classical indices and thresholds. The authors extrapolated these results to the entire elderly apneic population and suggested that the hypothetical causal link between OSA and dementia [4] may be explained by a “nonvascular” pathophysiological pathway involving cerebral accumulation (and/or defective elimination) of Tau and/or beta-amyloid proteins.

While this latter hypothesis is reinforced by pathophysiological studies showing the importance of sleep continuity, the role of slow wave sleep and the deleterious effects of variations in distal arteriolar pressure caused by respiratory effort in the efflux of these neurotoxic proteins [5], ruling out a microvascular hypothesis seems in our opinion somewhat premature.

In fact, several recent studies have clearly demonstrated the highly probable role of sleep-related obstructive respiratory pathology in accelerating the phenomena involved in the fragility of the cerebral white matter: volume of leukoaraiosis or presence/number of microbleeds such as the widening of the perivascular spaces, abnormal diameter of retinal arteries and or veins [6, 7]. Meta-analyses have their limitations [8, 9], but such methodological considerations are also present in randomized controlled studies such as the one we are discussing here. The absence of daytime symptoms associated with the nocturnal respiratory event, which are essentially desaturating events, the selection of volunteers (and not patients) who do not have significant cardiovascular comorbidities, the unusual methodology for performing diagnostic assessments of sleep-disordered breathing [10] focusing mainly on the load of oxyhemoglobin desaturations and thereby excluding patients with hypopneic syndromes that fragment their sleep but had no impact on nocturnal hypoxic load, implies that we are probably not talking about similar populations included in other clinical studies. Such “negative” results have been reported in elderly apneic patients according to brain volume morphometry analysis [11] and the importance of selection criteria at the inclusion has also been discussed as a major potential bias [12].

Then, the contradictory results of cross-sectional studies concerning microvascular alterations in the form of leukoaraiosis or distal microhemorrhagic lesions of the cerebral circulation should be explained by numerous factors. The first of these is represented by the confounding risk factors often present in the apneic population and a source of microvascular fragility: elderly [13], arterial hypertension [14], and diabetes [15]. It is therefore very difficult to distinguish between chronic intermittent hypoxic stress and/or repeated cyclical variations in distal cerebral arterial pressure and the consequences of hypertension at this arteriolar level, especially given the very high prevalence of hypertension in the clinical apneic population. Similarly, age is also a major factor in the increase in cerebral microvascular and arteriolar frailty, and the increased incidence of OSA at this age may be interpreted as a coincidence and/or as an additional aggravating factor for gray matter alterations [16].

There are still many other points to discuss after exploring these interesting clinical studies. The duration of exposure to chronic intermittent hypoxia undoubtedly has an important role to play but this is not taken into account, particularly in asymptomatic subjects. The conjunction of risk factors associated with microvascular remodeling and genetic susceptibility are also factors that can modify the phenotypic expression of this cerebral microvascular fragility [17]. In animal models presenting cerebral microvascular fragility (spontaneously hypertensive rats or “SHR”), it has been shown, for example, that long-term sleep fragmentation-induced pathological effects on the deep white matter which was not found in control Wistar Kyoto rats [18].

The specific effect of a sleep fragmentation “component” could thus have an essential role in this distal vascular fragility [19] and had not been taken into account in the study presented here. The authors claimed also in support of their findings of a lack of an association of OSA with cerebral small vessel disease (CSVD), found in a Mendelian randomization study reporting that genetically predicted OSA did not associate with any measure of CSVD, including white matter hyperintensities, in people of European background [20].

Moyamoya syndrome is a rare occlusive cerebrovascular arteriopathy with a significantly increased risk for stroke in childhood. Some cases have been published as an unusual but dramatic event after adenotonsillectomy (in a context of suspected or confirmed OSA in children) [21]. In young adults, the association between Moyamoya syndrome and sleep apnea is not well recognized. In this very particular context, thrombotic vasculopathy is also often associated with other cardiovascular risk factors and the specific impact of OSA is probably difficult to determine [22]. The role of autonomous nervous system alteration in the context of OSA on brain small vessels has also been evaluated. A recent study shows an important effect of nighttime heart rate variability on the association between the apnea plus hypopnea index and the cerebral small vessels disease score and provides further support for the role of sympathetic overactivity on this association [23].

Brain MRI morphometric analysis data on the volume of leukoaraiosis are now readily available and should help us to answer the important question of the cognitive consequences of such brain disorder. Then, a point raised in the course of OSA is the cognitive consequences of such microvascular fragility, with still a great deal of uncertainty as to the specific and independent impact of OSA and its various components on cognitive inflection. It is also interesting to note in the work published in SLEEP the absence of impact of sleep-related breathing disorder discovered in this selected population on cognition in almost of its domain and a recent paper from this Team confirmed these results in a larger subpopulation of the main study [24]. This is consistent with data from other selected elderly cohorts including superhealthy subjects [25] but differs from the results obtained in the younger general population [26]. Here again, the consequences of OSA on brain vascular fragility are probably age-dependent with a possible greater fragility of the youngest and most symptomatic patients (for some of them!) and possibly a protective effect of chronic intermittent hypoxia in some (other) patients as has been demonstrated experimentally by several authors [27] according to a potential preconditioning effect.

Finally, the absence of a “protective” impact of aspirin on this cerebral/retinal microvascular fragility in these contexts must be interpreted with particular attention because of the number of patients included that may be insufficient to conclude in this ancillary study.

Ultimately, the present negative results have the merit of calling into question pathophysiological elements that meta-analyses could have made us forget. The role of the phenotype of the apneic patient [28] returns to the “heart of the discussions” at the end of this work and a better knowledge of subjects at risk will be based as much on observational clinical databases “from real life” as on studies with a sophisticated design but representing only a limited fringe of the population of apneic patients seen in our sleep laboratories. This work should not discourage the realization of prospective longitudinal studies testing the favorable impact of the treatment of OSA on the upstream elements of vascular or nonvascular dementia. The stakes in public health and above all on an individual and personal level, of the prevention of these neurodegenerative diseases are far too important!

Disclosure Statement

Financial disclosure. All the authors declare no financial arrangements or connections to declare in relation to the submitted paper. Non-financial disclosure. All the authors declare no conflict of interest to declare in relation to the submitted paper.