1.1

Introduction

Prevalence of sleep apnea syndrome (SAS) in middle-aged subjects (from 30 to 60 years) is estimated at 2% in women and at 4% in men . SAS is currently considered as a public health problem on account of its elevated frequency and because of the forms of neuropsychic and cardiovascular morbidity for which it is held responsible ; it consequently necessitates screening and specific therapeutic management. From a vascular standpoint, SAS is a risk factor for high blood pressure (OR = 2.89 CI% [1.47–5.69]) , arrhythmia by atrial fibrillation (AAF) (OR= 4.02 CI% [1.03-15.74]) , cardiac insufficiency (OR = 1.68 CI% [1.02–2.76]) and stroke (OR = 2.44 CI% [1.03–5.80] et 3.48 CI% [1.31–9.23]) .

The etiological correspondences between SAS and stroke are numerous. First of all, apneas are likely to diminish partial pressure of oxygen in the blood, thereby provoking desaturation. As for secondary chronic hypoxemia, it could be responsible for blood pressure and the appearance of other cardiovascular disorders. More specifically, it can engender arrhythmias, cerebral hemodynamic changes or a prothrombotic state leading to strokes .

After a stroke, SAS prevalence is highly elevated, at times reaching 70% according to studies on patients after a recent stroke, and is greater than that of a population paired in terms of age and sex . Presence of SAS after a stroke is associated with higher one-year mortality , and some studies have objectified functional consequences. For example, Good et al. observed a less satisfactory score on the Barthel Index 3 to 12 months after a stroke in patients suffering from SAS, while Turkington et al. confirmed these results at six months. Cherkassky et al. likewise demonstrated less satisfactory functional recovery using the Functional Independence Measure (FIM); this was particularly the case in patients with low FIM scores on admission. For other authors (Iranzo), while respiratory sleep disorders were correlated with early neurological deterioration from the first to the third day using the Scandinavian Stroke Scale (SSS), they were not correlated to less satisfactory functional evolution using the Barthel Index . In spite of its sizable prevalence and demonstrated impact in the general population with regard to the cognitive functions (impaired attentiveness, reasoning, learning abilities, vigilance and memory) , the repercussions of SAS on neurological and functional recovery after stroke have not been widely studied.

Given the existence of treatment by continuous positive pressure (CPP) , screening is undeniably useful; above and beyond the expected functional consequences, it contributes to management of cardiovascular risk factors. However, it is often difficult to organize a polysomnography or sleep study, which is the reference examination. For this reason, acquisition of an oximetry apparatus such as ApneaLink, which screens respiratory sleep disorders by non-invasively recording pulse oximetry, respiratory flow and snoring, is an interesting option for physical medicine and rehabilitation (PMR) units treating recent stroke patients, who are particularly at risk. Evaluation of nocturnal oxygen saturation by ApneaLink seems highly reliable, with sensitivity at 100% and specificity at 97.5% for an apnea-hypopnea index (AHI) higher than 10 having been reported by some authors , and sensitivity at 91% and specificity at 95% having been obtained for AHI higher than 15 in 59 diabetic subjects with a mean age of 57 years .

SAS is a respiratory disorder defined by occurrence of 5 apnea or hypopnea events per hour of sleep (IAH > 5) and the appearance of nocturnal (snoring) and diurnal (sleepiness) symptoms. This definition of SAS was formalized in 1999 in the United States and more recently in France. Prior to that date, definitions were not nearly as precise. The cutoff point was 5, 10, 15, 20 and even 25 depending on the studies, and it varied according to the screening apparatus (respiratory polygraphy, polysomnography ventilator) and the definition of apnea and hypopnea that had been given.

The objective of the work reported in this paper was to analyze the possible impact of SAS on the neurological and functional evolution and the attentional abilities of the patients studied.

1.2

Patients et methods

The monocentric study was carried out in a neurologically oriented physical and rehabilitation medicine unit. It received the authorization of the CEERB (Ethics Committee) of the St Louis-Lariboisière-Fernand Widal Hospital Group.

The inclusion criteria were: more than 18 years old, non-opposition of the patient or relative and/or a close friend to his or her participation in the study in the event of aphasia, ischemic or hemorrhagic stroke less than six months before, and affiliation to a social security scheme.

The non-inclusion criteria were: SAS previous to the documented and treated stroke, cardio-respiratory insufficiency, dementia or inability to take the tests, rehabilitation having begun in another PRM unit.

All of the patients hospitalized in the unit and corresponding to the criteria of inclusion and exclusion were consecutively included. Inclusion was carried out during the first 10 days of hospitalization in the PRM unit. Patients were monitored over a period of two months.

The general data (age, sex, weight, height, body mass index (BMI), cardio-vascular risk factors, date of stroke and time elapsed since stroke, type of stroke – ischemic or hemorrhagic – and brain injury location) were taken from the patients’ records.

The clinical assessment tools were: score on the National Institute of Health Stroke Scale (NIHSS) for overall severity of the stroke, Fugl Meyer (FM) Score for motor deficiency, calculated by the investigators, and the Functional Independence Measure (FIM) for functional consequences, as determined by the entire team in a review meeting. The different assessments were carried out during the first 10 days following admission and at two months (± 15 days). A patient’s functional and neurological recovery was estimated in terms of Delta FM (= FM at 2 months – FM admission), Delta NIHSS (= NIHSS à 2 months – NIHSS admission) and Delta FIM (= FIM at 2 months – FIM admission).

As soon as the patients were found to be sufficiently alert, their attentional abilities were appraised by the speech therapists of the unit, using the “Battery Attention William Lennox” (BAWL) program , which consists in eight successive tests taken on a computer. As evaluation parameters, we chose two out of the eight tests, the first involving speed of execution, and the second concerning inhibitory and discriminatory capacity. The rapidity test measured visual reaction time; the subject had to react as quickly as possible to the appearance of a black square that was systematically presented in the center of the screen. The second test recorded simple binary reaction time; the subject was asked to press the “answer” button only when two signals (red cross and blue circle) had appeared, and not to react when two “distracter” signals (blue cross and red circle) were shown. Results were expressed in seconds and in number of omissions and “false alarms”.

SAS screening was carried out through an interview of the patient and his family using the Epworth Sleepiness Scale (ESS) questionnaire , which measures drowsiness in eight different situations, and also by assessing blood oxygen saturation using nocturnal oximetry (ApneaLink) during the first ten days of hospitalization so as to determine the apnea-hypopnea index (AHI). When this evaluation appeared to indicate SAS (AHI equal to or greater than 10), a polysomnography test was requested. Even though today’s American and French definition of SAS is set at IAH > 5, we chose a cutoff point of 10; as is reported in the review of the literature carried out Hermann and Bassetti , the majority of the studies on SAS and stroke have proceeded likewise, and our choice was particularly motivated by the characteristics of ApneaLink . Moreover, we defined SAS severity according to the AHI index, using the classification system validated by the Société française de pneumologie . SAS is considered as mild with AHI from 5 to 15, moderate from 15 to 30 and severe when higher than 30.

While hospitalized in the PRM unit, patients had access to the usual multi-daily rehabilitation program (kinesitherapy, occupational therapy and speech therapy) according to their needs.

1.3

Results

The 45 patients figuring in the protocol were prospectively included: 30 men and 15 women with a mean age of 60.9 years (SD = 11.5; 29–83). As concerns study of cardiovascular risk factors, mean BMC was 24, 15 patients smoked, 12 of them had high blood pressure, eight were diabetic, and 11 suffered from documented heart disease. On admission, average post-stroke time lapse was 26 days [3–101]. Ten strokes had been hemorrhagic, whereas the other 35 were ischemic. While 15 strokes involved the right side, 28 were left-sided, and the other two bilateral ( Table 1 ). There was no significant difference with regard to the existence or non-existence of SAS in the demographic data, except for age ( P = 0.004); the group suspected of SAS was pronouncedly older.

On the average, the BAWL tests were taken 28 days after admission to the unit [0–83] and 52 days after the stroke [7–109].

Two patients were lost to follow-up after having undergone initial examination and sleep recording, and one patient’s cardiovascular risk factors were not collected; that is why disease development analysis was finally limited to 43 patients.

1.4

Discussion

In this study, we sought out the possible repercussions of SAS on the neurological and functional development of patients having recently had a stroke; evolution of their attentional abilities was likewise assessed.

We did not discover convincing evidence of these repercussions, which have nonetheless been found by several authors ; a number of hypotheses have been put forward to explain the link. Cerebral hypoperfusion or blood pressure variability related to SAS could enlarge the penumbra area immediately after a stroke , or hematosis alteration secondary to SAS could modify cerebral blood, brain reorganization and, consequently, brain plasticity . Lastly, patients with SAS might present attentional deficits likely to complicate adaptation and acquisition of new aptitudes after a stroke .

The absence of a clear link between SAS and attentional or functional repercussions is perhaps largely due to the modes of evaluation applied. The FIM scale is no doubt lacking in sensitivity. In a previous study, it was used for assessment of the functional recovery on discharge of 30 patients . After adjustment for functional status on admission, nocturnal hypoxic events were found to be predictive of less satisfactory recovery (assessed in terms of FIM points), particularly in stroke patients with a low FIM score (< 70) on admission ( P = 0.025). It should also be noted that nocturnal respiratory disorders were observed in only 36.7% of the patients; moreover, stroke severity was not described. For all these reasons, it is difficult to draw a comparison with our groups of subjects.

Nor did we find a connection between existence of SAS and evolution of attentional abilities. Our admittedly arbitrary choice of two BAWL (battery attention) tests evaluating speed of execution and inhibitory and discriminatory capacities seems judicious. On the other hand, our choice of test-taking times may have been debatable. In point of fact, these moments were not standardized, the reason being that we had initially decided that the test was to be taken “as soon as possible”, that is to say once a patient was deemed able to take it. As a result, the tests were given at variable time intervals following admission, with considerable dispersion, and their heterogeneity may have entailed evaluation bias. It is also possible that test results varied according to time of day, especially in brain-damaged patients undergoing intensive rehabilitation and showing signs of fatigue. Variability of brain injury was yet another reason why the attentional tests were carried out at variable time intervals subsequent to the stroke. Last but not least, the assessment at two months may have taken place too early to underscore a difference, and the number of patients included was probably insufficient, which may have meant that the study was lacking in power. It would be interesting to complete this work by performing neuropsychological BAWL attention tests at two months, and associating them with a new ApneaLink recording in order to determine possible evolution in SAS prevalence according to neurological and functional recovery.

The consequences of SAS severity on neurological and functional evolution and attentional abilities have not been demonstrated. The hypothesis according to which severe SAS would be more deleterious than mild SAS has consequently not been confirmed. In upcoming study of the subject, it will be necessary to supplement classification of SAS severity with consideration of the degree of associated daytime sleepiness, which could be categorized, in accordance with existing recommendations for clinical practice , as mild, moderate or severe. It will also be necessary to take another look at the choice of the AHI index as a means of defining SAS. While our cutoff point of 10 was congruent with the literature, the notion of severity, particularly in mild SAS, is defined as equal to or greater than five apneas-hypopneas per hour of sleep. On another score, we have observed that the previous studies on SAS repercussions did not specify its severity; it was simply indicated that mean AHI ranged from 9.5 ± 9.6 in the study by Good et al. to 27.7 ± 26.6 in the study by Iranzo et al. ; other authors, such as Hui, determined a percentage of patients according to their AHI scores. However, the three sub-groups categorized by Hui (AHI > 10, 15 and 20) did not reflect SAS severity in agreement with the consensus definition.

Our study also underscored a high prevalence of SAS in a group of brain-damaged patients, of whom 62% presented respiratory disorders in the recording for screening carried out using the ApneaLink apparatus. These results are in agreement with the figures found in the literature, in which post-stroke SAS frequency ranges from 50 to 70% .

At another level, we encountered considerable difficulty in obtaining access to polysomnography laboratories in view of confirming the existence of nocturnal respiratory disorders. Early, relatively well-tolerated screening is that much more advantageous, especially insofar as the sensitivity and specificity of the ApneaLink apparatus are demonstrably good . This is illustrated by the fact that out of seven polysomnography studies carried out following positive screening, six cases of SAS were confirmed (Se = 85.7%). That much said, four of our patients were treated with nocturnal continuous positive pressure subsequent to this work on account of the above-mentioned difficulty of obtaining access to polysomnography laboratories and loss of follow-up for the relevant data.

It is also important to mention that use of this kind of nocturnal screening apparatus requires specific training for caregivers, including nurses and nursing assistants on the night shift in our PRM units. In our PRM unit, we have organized daytime and nighttime information and instruction sessions expressly addressed to the different caregiving teams.

Generally speaking, patient tolerance of the Apnea-Link system has been satisfactory. However, recording has often had to be redone because of insufficient allotted time or due to the difficulties encountered by caregivers using the system (poor management of the apparatus batteries or involuntary unplugging and disconnection…). A posteriori, we regret that our feasibility study was exclusively qualitative, and is consequently wanting in precision. Moreover, a substantial number of patients refused nocturnal screening out of fear that their precarious sleep would be altered by the application of nasal cannula and the recording device. The fear factor limited inclusions and may even have engendered selection bias. A possible limit to our study consisted in our inability to quantify the exact number of refusals.

1.5

Conclusion

No difference in recovery nor in intensity of attentional disorders has been shown to exist between patients presenting or not presenting more or less severe SAS after a stroke; that much said, our study is hardly exempt from biases and limits. We believe it would be of marked interest to pursue this work with studies featuring a schema that would be more rigorous in terms of chronology of the evaluations and, perhaps, of choice of tests to be given. Middle to long-term monitoring of apnea development according to evolution of sequels to the stroke would likewise be of interest.

Disclosure of interest

The authors declare that they have no conflicts of interest concerning this article.

2.1

Introduction

La prévalence du syndrome d’apnée du sommeil (SAS) chez les sujets d’âge moyen (de 30 à 60 ans) est estimée à 2 % chez les femmes et 4 % chez les hommes . Le SAS est considéré aujourd’hui comme un problème de santé publique du fait de sa fréquence élevée et des morbidités neuropsychiques et cardiovasculaires qui lui sont attribuées , justifiant ainsi un dépistage et une prise en charge thérapeutique spécifique. Sur le plan vasculaire, la présence d’un syndrome d’apnée du sommeil est un facteur de risque d’hypertension artérielle (OR = 2,89 IC % [1,47–5,69]) , d’arythmie par fibrillation auriculaire (ACFA) (OR = 4,02 IC % [1,03–15,74]) , d’insuffisance coronarienne (OR = 1,68 IC % [1,02–2,76]) et d’accident vasculaire cérébral (AVC) (OR = 2,44 IC % [1,03–5,80] et 3,48 IC % [1,31–9,23]) .

Les liens étiologiques avancés entre SAS et AVC sont multiples. Tout d’abord, les apnées diminueraient la pression partielle en oxygène dans le sang provoquant une désaturation. L’hypoxémie chronique secondaire serait elle-même responsable d’une hypertension artérielle et de l’apparition d’autres troubles cardio-vasculaires. En effet, le SAS engendrerait également des arythmies, des changements hémodynamiques cérébraux ou encore un état pro-thrombotique sanguin à l’origine d’accidents vasculaires cérébraux .

Après AVC, la prévalence du SAS est très élevée jusqu’à 70 % selon les études chez les patients après un accident vasculaire cérébral récent, et est supérieure à celle d’une population appariée en âge et en sexe . La présence d’un SAS après un AVC est associée à une plus grande mortalité à un an et quelques études objectivent un retentissement fonctionnel. Il en est ainsi de Good et al. qui observent un moins bon indice de Barthel à 3 et 12 mois après AVC, chez les patients avec un SAS et Turkington et al. confirment ces résultats à 6 mois. Cherkassky et al. montrent également une moins bonne récupération fonctionnelle par une évaluation par la mesure de l’indépendance fonctionnelle (MIF) et plus particulièrement chez les patients ayant une MIF faible à l’admission. Pour d’autres auteurs (Iranzo), si les troubles respiratoires du sommeil étaient corrélés à une détérioration neurologique précoce entre le 1 ^er et le 3 ^e jour évaluée sur la Scandinavian stroke scale (SSS), ils ne l’étaient pas à une moins bonne évolution fonctionnelle évaluée par l’indice de Barthel . Malgré sa prévalence importante et son retentissement démontré dans la population générale sur les fonctions cognitives: atteinte des capacités attentionnelles, de raisonnement, d’apprentissage, de vigilance et de mémoire , l’influence de la présence d’un SAS sur l’évolution neurologique et fonctionnelle après AVC a été peu étudiée.

Du fait de l’existence d’un traitement par pression positive continue (PPC) , le dépistage a une véritable utilité car au-delà des conséquences fonctionnelles attendues, il fait partie de la prise en charge indispensable des facteurs de risque cardio-vasculaires. Cependant l’obtention d’une polysomnographie, examen de référence, peut parfois être difficile. Pour cette raison, l’acquisition d’appareil d’oxymétrie type ApneaLink ^® qui permet d’enregistrer l’oxymétrie de pouls, le flux respiratoire et le ronflement de manière non invasive et donc de dépister des troubles respiratoires du sommeil est une option intéressante pour des unités de médecine physique et de réadaptation (MPR) prenant en charge des patients après accident vasculaire récent et donc dits « à risque ». La fiabilité du recueil de la saturation nocturne par ApneaLink ^® parait bonne avec une sensibilité de 100 % et une spécificité de 97,5 % pour un index apnée-hypopnée (IAH) supérieur à 10 rapportée par certains auteurs et une sensibilité de 91 % et une spécificité de 95 pour un IAH supérieur à 15 chez 59 sujets diabétiques âgés en moyenne de 57 ans .

Enfin, le syndrome d’apnée du sommeil est un trouble respiratoire défini par la survenue de plus de 5 apnées ou hypopnées par heure de sommeil (IAH > 5) et l’existence de symptômes nocturnes (le ronflement essentiellement) et diurnes (la somnolence principalement). Cette définition du SAS a été officialisée depuis 1999 aux États-Unis et de façon plus récente en France. Avant cela, les définitions étaient plus floues. Le cut-off était parfois à 5, 10, 15, 20 ou encore 25 selon les études et variait en fonction de l’appareil de dépistage (polygraphie, polysomnographie ventilatoire) et la définition de l’apnée et de l’hypopnée elles-mêmes.

L’étude rapportée ici avait pour objectif d’étudier l’éventuel retentissement du SAS sur l’évolution neurologique et fonctionnelle des patients et sur leurs capacités attentionnelles.

2.2

Patients et méthodes

Il s’agit d’une étude monocentrique effectuée dans un service de médecine physique et de réadaptation à orientation neurologique. Cette étude a reçu l’autorisation du CEERB du groupe St. Louis-Lariboisière-Fernand Widal.

Les critères d’inclusion étaient : patients de plus de 18 ans, non opposition du patient ou d’un membre de sa famille et/ou d’un proche à la participation à cette étude en cas d’aphasie, AVC ischémique ou hémorragique datant de moins de 6 mois et patient affilié à la sécurité sociale. Les critères de non inclusion étaient : SAS préalable à l’AVC documenté et traité, insuffisance cardio-respiratoire, démence ou incapacité à effectuer les tests, rééducation débutée dans un autre service de MPR. Tous les patients hospitalisés dans le service et répondant aux critères d’inclusion et d’exclusion étaient inclus de manière consécutive. L’inclusion était réalisée dans les 10 premiers jours de l’hospitalisation dans le service de MPR. Chaque patient était suivi sur une période de 2 mois.

Les données générales : âge, sexe, poids, taille, indice de masse corporelle (IMC), facteurs de risque cardio-vasculaires, date de l’AVC et délai depuis l’AVC, type d’AVC : ischémique ou hémorragique et localisation de la lésion cérébrale, étaient recueillies à partir du dossier du patient.

Les outils d’évaluation clinique étaient : score du National Institute of Health Stroke Scale (NIHSS) pour la sévérité globale de l’AVC, le score de Fugl Meyer (FM) pour la déficience motrice, réalisés par les investigateurs et la mesure d’indépendance fonctionnelle (MIF) pour le retentissement fonctionnel, déterminée par l’ensemble de l’équipe en réunion de synthèse. Ces évaluations étaient réalisées dans les 10 premiers jours après l’entrée des patients puis à deux mois (± 15 j). La récupération fonctionnelle et neurologique des patients étaient représentées par les Delta FM (= FM à 2 mois – FM admission), Delta NIHSS (= NIHSS à 2 mois – NIHSS admission) et Delta MIF (= MIF à 2 mois – MIF admission).

Dès que l’état de vigilance des patients le permettait, leurs capacités attentionnelles étaient évaluées par les orthophonistes du service, au moyen de la batterie attentionnelle de William Lennox (BAWL) . Il s’agit d’une succession de huit tests passés sur ordinateur. Nous avions choisi comme paramètres d’évaluation deux épreuves parmi les huit, la première explorant la rapidité d’exécution, la seconde explorant la capacité d’inhibition et de discrimination. La première étudiait le temps de réaction visuel centré : le sujet devait réagir le plus rapidement possible à l’apparition d’un carré noir présenté systématiquement au centre de l’écran. La seconde étudiait le temps de réaction simple binaire : le sujet ne devait actionner le bouton réponse qu’à l’apparition de deux signaux (croix rouge et rond bleu) en s’abstenant de réagir à la présentation de signaux distracteurs (croix bleue et rond rouge). Les résultats étaient exprimés en secondes, en nombre d’omissions et de fausses alertes.

Le dépistage du SAS était effectué à l’aide de l’interrogatoire du patient et de la famille par la réalisation de l’échelle d’Epworth ^© , questionnaire mesurant la somnolence diurne dans huit situations différentes, et par le recueil de la saturation nocturne par un appareil d’oxymétrie (Apnea Link ^® ) dans les dix premiers jours d’hospitalisation permettant le recueil de l’index d’apnée-hypopnée (IAH). Si cet examen était en faveur d’un SAS (IAH supérieur ou égal à 10), une polysomnographie était demandée. Malgré la définition actuelle américaine et française du SAS avec un IAH > 5, nous avons choisi un cut-off à 10 inclus car comme le rapporte la revue de littérature de Hermann et Bassetti , la majorité des études ayant étudié SAS et AVC ont utilisé cette limite et en raison des caractéristiques de l’ApneaLink . Nous avons également défini la sévérité du SAS des patients en fonction de l’index IAH, en utilisant partiellement la classification validée par la Société française de pneumologie . Un SAS est dit léger pour un IAH entre 5 et 15, modéré entre 15 et 30 et enfin sévère si supérieur à 30.

Au cours de l’hospitalisation dans le service de MPR, les patients bénéficiaient du programme pluri-quotidien de rééducation habituelle (kinésithérapie, ergothérapie et orthophonie) en fonction de leurs besoins.

2.3

Résultats

Les 45 patients prévus par le protocole ont été inclus de manière prospective : 30 hommes et 15 femmes avec un âge moyen de 60,9 (DS = 11,5 ; 29–83). Concernant l’étude des facteurs de risque cardio-vasculaire, l’IMC moyen des patients était de 24, 15 patients fumaient, 32 avaient une hypertension artérielle, 8 étaient diabétiques et 11 avaient une cardiopathie connue. Le délai post-AVC moyen à l’admission était de 26 jours [3–101]. Dix AVC étaient de cause hémorragique et 35 ischémique. Quinze AVC concernaient l’hémisphère droit, 28 l’hémisphère gauche et 2 AVC étaient bilatéraux ( Tableau 1 ). Il n’y avait pas de différence significative selon l’existence ou non d’un SAS concernant les données démographiques sauf pour l’âge ( p = 0,004) : le groupe suspecté de SAS étant plus âgé.

Tableau 1

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Enhancing Annalsvalue through quality Detection of progressive idiopathic scoliosis during growth using back surface topography: A prospective study of 100 patients Cognitive and behavioural post-traumatic impairments: What is the specificity of a brain injury ? A study within the ESPARR cohort Quality of life and satisfaction after multilevel surgery in cerebral palsy: Confronting the experience of children and their parents Communication disorders and executive function impairment after severe traumatic brain injury: An exploratory study using the GALI (a grid for linguistic analysis of free conversational interchange) The motor function measure to study limitation of activity in children and adults with Charcot-Marie-Tooth disease

Stay updated, free articles. Join our Telegram channel

Join