Benzodiazepines, CHRISTIAN GUILLEMINAULT,
M.D. Stanford,
Breathing,
and Sleep
ca/ifornia
The benzodiazepines are sedative hypnotic drugs, i.e., central nervous system depressant drugs, that may adversely affect the control of ventilation during sleep. Prescription of these drugs may worsen sleep-related breathing disorders, especially in patients with chronic obstructive pulmonary disease or cardiac failure. The most frequent users of sedative hypnotics are the polymorbid elderly with a secondary complaint of insomnia. Although the benzodiazepines may reduce sleep fragmentation, their long-term use may also cause health problems, such as complete obstructive sleep apnea in heavy snorers or short repetitive central sleep apnea in patients with recent myocardial infarction. Since drugs of this class vary in their effects, it is crucial to note the action of a given benzodiazepine on the control of vital functions during sleep.
he prevalence of sleep apnea in the general popuT lation of adult men 65 years of age and younger has been estimated to be between 1.5 and 5 percent. The frequency of sleep apnea increases with aging; men predominate among patients less than 50 years old, and postmenopausal women predominate in the over-50 group. Ancoli-Israel et al [l] have recently shown that an apnea index (number of apneas per hour of sleep) above 5 may correlate with decreased longevity in elderly women. Apnea is associated with sleep-related complaints such as daytime tiredness, fatigue, and sleepiness; nocturnal sleep disruption; and insomnia. BREATHING DISORDERS DURING SLEEP Chronic obstructive pulmonary disease (COPD) is the most common problem seen by pulmonary specialists, and impaired quality of sleep is frequently reported by patients. In 1982, Fleetham et al [2] objectively demonstrated the disruption of nocturnal sleep in patients with hypoxemic COPD due to the frequent arousals caused by respiratory stimuli, an observation corroborated by Calverley et al [3]. Sleep disruption may also be increased by xanthine derivatives, widely prescribed on a continuing basis in COPD patients. The insomnia reported by these patients is not significantly helped by oxygen administration through nasal prongs. The underlying mechanisms responsible for the repetitive nocturnal sleep disruptions are not well elucidated. Sleep-related changes in cough capability, ciliary movements, and bronchial excretions have been cited as possible factors in the repetitive awakenings. The sleep fragmentation observed in COPD has also been hypothesized as playing a role in the blunting of hypoxic and hypercapnic responses that is sometimes observed in these patients. Obesity is commonly associated with obstructive sleep apnea (OSA). In one large series of patients with obstructive sleep apnea syndrome seen successively, two thirds had a body mass index above 28. Obesity may also be seen with COPD or in isolation, may be responsible for very significant chest-bellows disease, and may be associated with rapid eye movement (REM)-sleep-related obesity-hypoventilation, with repetitive awakenings from REM sleep. HYPNOTICS AND BREATHING DURING SLEEP Oxygen Saturation and Hemodynamics
From the Stanford University Medical Center, Stanford, California. This study was supported by grant AGO7772 from the National institute of Aging. Requests for reprints should be addressed to ChristIan Guillemlnault, M.D., Stanford Sleep Disorders Center, 701 Welch Road, Suite 2226, Palo Alto, California 94304.
The breathing disorders cited earlier lead to chronic complaints of poor nocturnal sleep, for which hypnotic drugs are frequently prescribed on a long-term basis. However, hypnotic drugs, including the benzodiazepines, may depress the central nervous system (CNS) and adversely affect the controls of ventilation during sleep, with worsening of the sleep-related breathing disorder. In 1972, Gaddie et al [4] showed that the benzodiazepines may cause hypoventilation in patients with severe COPD, due to the decreased ventil-
March 2, 1990
The American Journal of Medicine
Volume 88 (suppl 3A)
3A-25s
SYMPOSIUM
ON INSOMNIA / GUILLEMINAULT
atory response to carbon dioxide, as investigated by Geddes et al [5] and Rudolf et al [6]. There may, however, be important differences among benzodiazepines, as outlined by Cohn [7] in 1983. Clark et al [8] and Model [9] reported serious benzodiazepineinduced respiratory depression in patients with COPD, although Cummiskey et al [lo] and Midgren et al [ill noted that these drugs may improve sleep and sleep structure. Because of the depressant effect of certain drugs on the central nervous system, caution should be exercised in prescribing such medications to patients with obstructive or restrictive lung disease whose decreased respiratory effort causes a decrease in oxygen tension with resulting impact on oxygen saturation. Adverse effects on ventilation during sleep may be unrelated to drugs, however, and may be the result of long-standing physical causes such as COPD, marked abdominal obesity, or OSA. During sleep, oxygen saturation decreases in patients with these conditions. The degree of decrease in oxygen saturation seems to be directly related to the amount of oxygen saturation during wakefulness. However, arterial blood gas measurements must be obtained on a subject kept supine for 20 minutes prior to blood drawing to approximate the sleep-related risk of decrease in oxygen. Systematic polygraphic recordings have indicated, however, that values calculated during wakefulness do not accurately predict the lowest oxygen saturation during sleep, and hypoxemia may easily be measured during sleep with noninvasive oximeters. The significant variable with respect to oxygen saturation, alveolar hypoxia, depends upon the severity of individual polymorbidity. It is related to the degree of vasoconstriction in the pulmonary circulation and is responsible for transient or long-term pulmonary hypertension. Transient hypoxemia, as seen in repetitive OSA, may not lead to significant alveolar hypoxia if each apneic event is short-lived and if arousalrelated hyperventilation rapidly normalizes oxygen saturation levels. However, mechanical factors related to a significant increase in intrathoracic pressure, as demonstrated by increasingly negative endoesophageal pressure with each inspiratory trial during obstructive apnea, may also play a role in transient hemodynamic changes. In a patient presenting with chronic wake hypoxemia and blood gas measurements indicating oxygen saturation located on the steep portion of the hemoglobin-oxy-hemoglobin curve, a moderate decrease in oxygen tension can cause a much more significant decrease in oxygen saturation. Besides producing vasoconstriction of the pulmonary circulation, alveolar hypoxia results in cardiac arrhythmias, particularly ventricular arrhythmias. Arrhythmias are more frequent during sleep than during wakefulness due to more pronounced changes in the blood gases during sleep and to the sleep- and sleep state-related changes in autonomic nervous system controls of the cardiovascular system.
activation of many upper airway reflexes, the purpose of which is to enlarge the airway, despite a progressive increase in negative intrathoracic pressure that peaks at end inspiration. Negative intrathoracic pressure translates into negative transpharyngeal pressure, which, if unopposed, leads to a collapse of the noncartilagmous part of the upper airway. Also, the supine position of the subject during sleep exposes the diaphragm, abdomen, jaw, and related structures to atmospheric pressure in a different manner than during wakefulness. The position of the patient while asleep and sleep per se have an impact on the gamma loop of many skeletal muscles. This usually results in a greater laxity of the pharyngeal muscles. With REM sleep, an active inhibition of muscle tone occurs that involves a pathway from the locus ceruleus to the inhibitory medullary reticular formation of Magoun and Rhines and a descending reticulospinal pathway with relay in a spinal interneuron and terminals on the spinal motor neuron. Activation of this pathway leads to a hyperpolarization of the spinal motor neuron and a decrease in muscle tone. The genioglossus and geniohyoid muscles undergo such a decrease in tone during REM sleep, and CNS depressant drugs appear to affect the contraction of these muscles during sleep. Alcohol, which depresses the CNS, significantly decreases the activity of the genioglossus muscle during sleep and may be a factor in snoring (partial obstruction of the upper airway) or OSA (complete upper airway obstruction). Apnea leads to hypoxemia, but as already mentioned, partial or complete obstruction, with persistence of diaphragmatic movement., also induces significant changes in the intrathoracic pressure that can be documented by simultaneous monitoring of esophageal pressure or central venous pressure. A Mtiller maneuver is performed with each apnea, if muscle coordination is also impaired during expiration [12]. An active expiration will occur with a Valsalva maneuver (expiratory effort against a closed glottis). These maneuvers impact mechanically on ventricular load and ventricular ejection. Other changes in the cardiovascular system are also mediated through stimulation of the autonomic nervous system. Classically, the Valsalva maneuver consists of four phases, but the “Valsalva ratio” is commonly used: the tachycardia resulting from the reduced venous return to the right atrium and falling cardiac output (phase 2) is compared with the bradycardia resulting from the overshoot in blood pressure (phase 4) [13]. These maneuvers are performed during sleep, when the autonomic nervous system balance is different from that during wakefulness. Different baselines are present during non-rapid eye movement and REM sleep, and the Mtiller and Valsalva maneuvers are thus associated with a more pronounced vagal activity than during wakefulness. These differences in state-related autonomic nervous system settings must be considered when analyzing the impact of hypoxia, mechanical factors, and drugs such as benzodiazepines on the cardiovascular system [141.
Upper Airway Muscles, Miiller and Valsalva Maneuvers
Hypnotics, including the benzodiazepines, influence the controls of the upper airway muscles. Breathing is a coordinated act, with strict regional organization. Inspiration (diaphragmatic movement) is preceded by
3A-26s
March 2, 1990
The American Journal of Medicine
Snoring, Hypnotics, and the Elderly
The significance of decreased coordination between diaphragmatic and upper airway muscles during sleep after benzodiazepine intake is poorly documented.
Volume 88 (suppl 3A)
SYMPOSIUM
However, regular snorers (i.e., subjects presenting with only partial airway obstruction) given flurazepam 30 mg at bedtime have OSA during the drug night [15]. In these patients, flurazepam had the same effect as alcohol, i.e., a complete airway obstruction was noted compared with baseline. Considering the body of knowledge currently available on sleeprelated cardiovascular changes associated with OSA and on the interaction between the control of breathing and sedative hypnotics, chronic snoring should be systematically investigated before prescription of a sedative hypnotic, since a partial obstruction might thereby be changed into a complete one. Unfortunately, those most at risk of presenting with OSAthe elderly and, especially? postmenopausal womenare also those who chromeally use these drugs. Elderly subjects, in general, are more likely to present with morbidity; health problems are associated with poor sleep at any age, and this is particularly true when chronic insomnia is found in the elderly. Iatrogenically induced OSA encourages the development of sleep-related cardiac arrhythmias in the elderly. We investigated healthy elderly subjects who had been deprived of sleep on the preceding night or had been given 0.6 mglkg whiskey one hour prior to bedtime. These two conditions were selected because of their depressant effect on control of breathing during sleep. Elderly subjects with a respiratory disturbance index between 5 and 10 (i.e., experiencing five to 10 apneashypopneas per hour of sleep) had a clear increase in the number of sleep-related apneas, and in the case of one patient, salvos of premature ventricular complexes occurred in association with the environmentally induced increased sleep apneas [161. HYPNOTICS AND CORONARY ARTERY DISEASE The direct interaction between sedative hypnotic prescription and cardiovascular functioning during sleep has been little investigated to date. In a study of the effect of flurazepam, triazolam, and temazepam on sleep in middle-aged adults complaining of insomnia but without known cardiac lesions, we have noted no significant cardiovascular problems. However, in clinical practice, sedative hypnotics, including the benzodiazepines, are frequently prescribed for patients with cardiovascular diseases, particularly coronary heart disease (angina, myocardial infarction) and for patients with moderate symptoms of disease who frequently present with a disturbed nocturnal sleep and, at times, sleep-onset insomnia. The influence of benzodiazepines on cardiac function in patients with disturbed sleep can be seen particularly in the development of repetitive apneas during sleep. The occurrence of short apneas would seem to result from purely hemodynamic causes. However, there is some evidence that appears to implicate benzodiazepine hypnotic drugs in their development and continuation. Nine patients treated for myocardial infarction and receiving a benzodiazepine at bedtime during the previous three to six months were monitored polygraphitally during sleep at three different times: at entry, while taking the drug; on the last night of a three-day withdrawal from benzodiazepine; and on the third night after resumption of their prescribed regimen. At baseline with benzodiazepine, mean total sleep
ON INSOMNIA/GUlLLEMlNAULT
time was 342 i 32 minutes, indicating a poor quality of sleep despite hypnotic intake. The mean percentage of REM sleep was 15 +- 3.8 percent. Central apneas of between 10 and 18 seconds’ duration were seen, predominantly during REM sleep, in three patients. The number of central apneas varied between 32 and 151. Two of the three patients presented clusters of short central apneas. During the withdrawal period, patients’ sleep was worse on the third drug-free night: mean total sleep time was 317 2 47 minutes. However, none of the patients studied at this time presented with apneas. When patients resumed benzodiazepine intake, abnormal breathing patterns were again noted in the same three patients as at baseline. The lowest oxygen saturation during the night always appeared in association with the pattern of repetitive central apneas under baseline and drug conditions (mean +- SD were as follows: baseline oxygen saturation, 85 i: 6 percent; off-drug oxygen saturation, 90 +5 percent; on-drug oxygen saturation, 83 t 5.3 percent). The study showed that a breathing pattern, which occasionally led to a brief episode of oxygen desaturation, was associated with benzodiazepine intake, at least in three post-myocardial infarction patients. There is a time lag between the return of oxygenated blood from the right circulation to the left atrium and ventricle (i.e., into the coronary arteries) and that moment when peripheral and central chemoreceptors respond to changes in the blood oxygen content. This delay, short in normal subjects, increases significantly with left ventricular failure and is at least partially responsible for repetitive central apneas or for the development of classic Cheyne-Stokes respiration. However, this mechanism alone does not explain the short, repetitive apneic events seen in some of our patients since the short apneas disappeared on benzodiazepine withdrawal. In persons with existing cardiac lesions, benzodiazepines may also lead to repetitive short central apneas accompanied by brief, more marked, decreases in oxygen saturation. The combination of benzodiazepine and cardiac lesion seems to have an additive effect in certain patients, independent of the mechanisms involved. The effect of these decreases in oxygen saturation on an already compromised heart is still unclear, as is the impact of benzodiazepines on baroreceptors, particularly in populations of at-risk subjects. However, this type of investigation into the impact of benzodiazepine sedation, as well as information on benzodiazepines’ potential impact in patients presenting with cardiac failure, would be helpful to internists and to general practitioners, who face therapeutic dilemmas when treating these cases. COMMENTS Sleep and sleep states are associated with controls of vital functions that differ from those observed during wakefulness. The benzodiazepines, like any psychoactive drugs, have an impact on these controls. Recently, sleep researchers have given attention to the changes induced by CNS-depressant drugs on the control of air exchange during sleep, but little or no information is available on the impact of these drugs on the control of other vital functions during sleep. Nevertheless, epidemiologic studies in the Western
March 2, 1990
The American Journal of Medicine
Volume 88 (suppl 3A)
3A-27s
SYMPOSIUM
ON INSOMNIA I GUILLEMINAULT
Hemisphere indicate that the largest group of longterm consumers of sedative hypnotic drugs is the elderly. Elderly subjects often present with polymorbidity and, undoubtedly, breathing disorders and cardiovascular disease are more common in older age groups; sleep-related worsening of symptoms (due to sleep-state physiology) may lead to disturbed sleep and complaints of insomnia. To avoid biased data (including data that are difficult to interpret), many sleep laboratories study highly selected groups that exclude polymorbidity. These studies are useful; they assess the efficacy of sedative hypnotic drugs m sleep disturbance on welldefined populations of “healthy elderly” presenting only with sleep complaints. Very little investigation, however, has been carried out in populations presenting with ,several health problems, and little has been directed toward the impact of drugs on central controls of vital functions not during wakefulness but during sleep (or night), the time of peak hypnotic drug levels in the blood. COPD patients, cardiac patients, and obese patients assuredly are to be found among the ranks of chronic hypnotic users, who often combine alcohol with sedative intake. In a time of an increasingly aged population, nursing homes are more and more in demand, and families and guardians desire that the elderly sleep well at night without risking a confusional awakening that could lead to abnormal behavior endangering the resident as well as others. Epidemiologic studies also indicate that subjects in nursing homes are in poorer health than other elderly persons and are also long-term users of sedative hypnotic drugs [l]. In a very recent international symposium on sleep and health risks held in Marburg, West Germany, Kripke and Ancoli-Israel [17] reported on a large study performed in San Diego, California, and showed that elderly women who presented with an apnea index above 5 not only had a significantly greater chance of dying at an earlier age than others in their group, but also of dying at night while in bed
3A-28s
March 2, 1990
The American Journal of Medicine
(supposedly asleep). They also received more sedative hypnotic medications. Further investigation of the interaction of sedative hypnotics on the control of vital functions during sleep in health and disease, particularly in the older age groups, is clearly needed. REFERENCES 1. Ancoli-Israel 5. Klauber MR. Krioke DF. Parker L: Increased risk of mottalltv with sleeo apnea in nursing home patients:’ preliminary report (abstr). Sleep Res 1988; 17: 14i. 2. Fleetham J, West P, Mezon B, Conway W, Roth T, Ktyger M: Sleep, arousals, and oxygen desaturation in chronic obstructive pulmonary disease. Am Rev Respir Dis 1982; 126: 429-433. 3. Calverley PMA, Brezinova V, Douglas NJ, et at The effect of oxygenation on sleep quality in chronic bronchitis and emphysema. Am Rev Respir Dis 1982; 126: 206-210. 4. Gaddie J, Legge JS, Palmer KNV, Petrie JC, Wood RA: Effect of nitrazepam in chronic obstructive bronchitis. Br Med J 1972: 2: 688-689. 5. Geddes DM, Rudolf M, Saunders i(B: Effect of nitrazepam and flurazepam on the ventilatory response to carbon dioxide. Thorax 1976; 31: 548-551. 6. Rudolf M, Geddes DM, Turner JAM, Saunders KB: Depression of central respiratory drive bv nitrazeoam. Thorax 1978: 33: 97-100. 7. Cohn MA: Hybnotics and the control of breathing: a review. Br J Ciin Pharmacol 1983; 16 (suppl 2): 245%2505. 8. ClarkTJH, Collins JV, Tong D: Respiratory depression caused by nitrazepam In patients with respiratory failure. Lancet 1971; II: 737-738. 9. Model DG: Nltrazepam induced respiratory depression in chronic obstructive lung disease. Br J Dis Chest 1973; 67: 128-130. 10. Cummiskey J, Guilleminault C, Del Rio G, Silvestri R: The effects of flurazepam on sleep studies in patients with chronic obstructive pulmonary disease. Chest 1983; 84: 143-147. 11. Midgren B, Hansson L, Skeidsvoll H, Elmqvist D: The effect of nitrazepam and flunitrazepam on oxygen desaturation during sleep in patients with stable hypoxemic non-hypercapnic COPD. Chest 1989; 95: 765-768. 12. Guilleminault C, Dement WC (eds): Sleep apnea syndromes. New York: Alan R. Liss, 1978. 13. Ewing DJ: Practical bedslde investigation of diabetic autonomic failure. In: Bannister R, ed. Autonomic failure. London: Oxford University Press, 1983; 371-405. 14. Mancia G, Zanchett A: Cardiovascular regulation during sleep. In: Orem J, Barnes CD, eds. Physiology in sleep. New York Academic Press Inc., 1980; 2-55. 15. Guilleminault C. Cummiskev J. Silvestri R: Benzodiazeoines and resoiration during sleep. In: Udsin E, ‘Clarke P, Tellman D, Greenblatt D, Paul SM, eds. Pharmacology 0; benzodiazepines. London: Macmillan, 1982; 229-236. 16. Guilleminault C, Silvestri R, Mondini S, Coburn S: Aging and sleep apnea: action of benzodiazepine, acetazolamide, alcohol and sleep deprivation in a healthy elderly group. J Gerontol 1984; 39: 655-661. 17. Kripke D, Ancoli-Israel S: Health risk of insomnia (abstr). In: Peter JH, Podzus T, von Wicheri P, eds. Sleep and health risk, an international symposium. Marburg: Phillips Universitat Press, 1989; 90.
Volume 88 (suppl 3A)
M.D. Stanford,
Breathing,
and Sleep
ca/ifornia
The benzodiazepines are sedative hypnotic drugs, i.e., central nervous system depressant drugs, that may adversely affect the control of ventilation during sleep. Prescription of these drugs may worsen sleep-related breathing disorders, especially in patients with chronic obstructive pulmonary disease or cardiac failure. The most frequent users of sedative hypnotics are the polymorbid elderly with a secondary complaint of insomnia. Although the benzodiazepines may reduce sleep fragmentation, their long-term use may also cause health problems, such as complete obstructive sleep apnea in heavy snorers or short repetitive central sleep apnea in patients with recent myocardial infarction. Since drugs of this class vary in their effects, it is crucial to note the action of a given benzodiazepine on the control of vital functions during sleep.
he prevalence of sleep apnea in the general popuT lation of adult men 65 years of age and younger has been estimated to be between 1.5 and 5 percent. The frequency of sleep apnea increases with aging; men predominate among patients less than 50 years old, and postmenopausal women predominate in the over-50 group. Ancoli-Israel et al [l] have recently shown that an apnea index (number of apneas per hour of sleep) above 5 may correlate with decreased longevity in elderly women. Apnea is associated with sleep-related complaints such as daytime tiredness, fatigue, and sleepiness; nocturnal sleep disruption; and insomnia. BREATHING DISORDERS DURING SLEEP Chronic obstructive pulmonary disease (COPD) is the most common problem seen by pulmonary specialists, and impaired quality of sleep is frequently reported by patients. In 1982, Fleetham et al [2] objectively demonstrated the disruption of nocturnal sleep in patients with hypoxemic COPD due to the frequent arousals caused by respiratory stimuli, an observation corroborated by Calverley et al [3]. Sleep disruption may also be increased by xanthine derivatives, widely prescribed on a continuing basis in COPD patients. The insomnia reported by these patients is not significantly helped by oxygen administration through nasal prongs. The underlying mechanisms responsible for the repetitive nocturnal sleep disruptions are not well elucidated. Sleep-related changes in cough capability, ciliary movements, and bronchial excretions have been cited as possible factors in the repetitive awakenings. The sleep fragmentation observed in COPD has also been hypothesized as playing a role in the blunting of hypoxic and hypercapnic responses that is sometimes observed in these patients. Obesity is commonly associated with obstructive sleep apnea (OSA). In one large series of patients with obstructive sleep apnea syndrome seen successively, two thirds had a body mass index above 28. Obesity may also be seen with COPD or in isolation, may be responsible for very significant chest-bellows disease, and may be associated with rapid eye movement (REM)-sleep-related obesity-hypoventilation, with repetitive awakenings from REM sleep. HYPNOTICS AND BREATHING DURING SLEEP Oxygen Saturation and Hemodynamics
From the Stanford University Medical Center, Stanford, California. This study was supported by grant AGO7772 from the National institute of Aging. Requests for reprints should be addressed to ChristIan Guillemlnault, M.D., Stanford Sleep Disorders Center, 701 Welch Road, Suite 2226, Palo Alto, California 94304.
The breathing disorders cited earlier lead to chronic complaints of poor nocturnal sleep, for which hypnotic drugs are frequently prescribed on a long-term basis. However, hypnotic drugs, including the benzodiazepines, may depress the central nervous system (CNS) and adversely affect the controls of ventilation during sleep, with worsening of the sleep-related breathing disorder. In 1972, Gaddie et al [4] showed that the benzodiazepines may cause hypoventilation in patients with severe COPD, due to the decreased ventil-
March 2, 1990
The American Journal of Medicine
Volume 88 (suppl 3A)
3A-25s
SYMPOSIUM
ON INSOMNIA / GUILLEMINAULT
atory response to carbon dioxide, as investigated by Geddes et al [5] and Rudolf et al [6]. There may, however, be important differences among benzodiazepines, as outlined by Cohn [7] in 1983. Clark et al [8] and Model [9] reported serious benzodiazepineinduced respiratory depression in patients with COPD, although Cummiskey et al [lo] and Midgren et al [ill noted that these drugs may improve sleep and sleep structure. Because of the depressant effect of certain drugs on the central nervous system, caution should be exercised in prescribing such medications to patients with obstructive or restrictive lung disease whose decreased respiratory effort causes a decrease in oxygen tension with resulting impact on oxygen saturation. Adverse effects on ventilation during sleep may be unrelated to drugs, however, and may be the result of long-standing physical causes such as COPD, marked abdominal obesity, or OSA. During sleep, oxygen saturation decreases in patients with these conditions. The degree of decrease in oxygen saturation seems to be directly related to the amount of oxygen saturation during wakefulness. However, arterial blood gas measurements must be obtained on a subject kept supine for 20 minutes prior to blood drawing to approximate the sleep-related risk of decrease in oxygen. Systematic polygraphic recordings have indicated, however, that values calculated during wakefulness do not accurately predict the lowest oxygen saturation during sleep, and hypoxemia may easily be measured during sleep with noninvasive oximeters. The significant variable with respect to oxygen saturation, alveolar hypoxia, depends upon the severity of individual polymorbidity. It is related to the degree of vasoconstriction in the pulmonary circulation and is responsible for transient or long-term pulmonary hypertension. Transient hypoxemia, as seen in repetitive OSA, may not lead to significant alveolar hypoxia if each apneic event is short-lived and if arousalrelated hyperventilation rapidly normalizes oxygen saturation levels. However, mechanical factors related to a significant increase in intrathoracic pressure, as demonstrated by increasingly negative endoesophageal pressure with each inspiratory trial during obstructive apnea, may also play a role in transient hemodynamic changes. In a patient presenting with chronic wake hypoxemia and blood gas measurements indicating oxygen saturation located on the steep portion of the hemoglobin-oxy-hemoglobin curve, a moderate decrease in oxygen tension can cause a much more significant decrease in oxygen saturation. Besides producing vasoconstriction of the pulmonary circulation, alveolar hypoxia results in cardiac arrhythmias, particularly ventricular arrhythmias. Arrhythmias are more frequent during sleep than during wakefulness due to more pronounced changes in the blood gases during sleep and to the sleep- and sleep state-related changes in autonomic nervous system controls of the cardiovascular system.
activation of many upper airway reflexes, the purpose of which is to enlarge the airway, despite a progressive increase in negative intrathoracic pressure that peaks at end inspiration. Negative intrathoracic pressure translates into negative transpharyngeal pressure, which, if unopposed, leads to a collapse of the noncartilagmous part of the upper airway. Also, the supine position of the subject during sleep exposes the diaphragm, abdomen, jaw, and related structures to atmospheric pressure in a different manner than during wakefulness. The position of the patient while asleep and sleep per se have an impact on the gamma loop of many skeletal muscles. This usually results in a greater laxity of the pharyngeal muscles. With REM sleep, an active inhibition of muscle tone occurs that involves a pathway from the locus ceruleus to the inhibitory medullary reticular formation of Magoun and Rhines and a descending reticulospinal pathway with relay in a spinal interneuron and terminals on the spinal motor neuron. Activation of this pathway leads to a hyperpolarization of the spinal motor neuron and a decrease in muscle tone. The genioglossus and geniohyoid muscles undergo such a decrease in tone during REM sleep, and CNS depressant drugs appear to affect the contraction of these muscles during sleep. Alcohol, which depresses the CNS, significantly decreases the activity of the genioglossus muscle during sleep and may be a factor in snoring (partial obstruction of the upper airway) or OSA (complete upper airway obstruction). Apnea leads to hypoxemia, but as already mentioned, partial or complete obstruction, with persistence of diaphragmatic movement., also induces significant changes in the intrathoracic pressure that can be documented by simultaneous monitoring of esophageal pressure or central venous pressure. A Mtiller maneuver is performed with each apnea, if muscle coordination is also impaired during expiration [12]. An active expiration will occur with a Valsalva maneuver (expiratory effort against a closed glottis). These maneuvers impact mechanically on ventricular load and ventricular ejection. Other changes in the cardiovascular system are also mediated through stimulation of the autonomic nervous system. Classically, the Valsalva maneuver consists of four phases, but the “Valsalva ratio” is commonly used: the tachycardia resulting from the reduced venous return to the right atrium and falling cardiac output (phase 2) is compared with the bradycardia resulting from the overshoot in blood pressure (phase 4) [13]. These maneuvers are performed during sleep, when the autonomic nervous system balance is different from that during wakefulness. Different baselines are present during non-rapid eye movement and REM sleep, and the Mtiller and Valsalva maneuvers are thus associated with a more pronounced vagal activity than during wakefulness. These differences in state-related autonomic nervous system settings must be considered when analyzing the impact of hypoxia, mechanical factors, and drugs such as benzodiazepines on the cardiovascular system [141.
Upper Airway Muscles, Miiller and Valsalva Maneuvers
Hypnotics, including the benzodiazepines, influence the controls of the upper airway muscles. Breathing is a coordinated act, with strict regional organization. Inspiration (diaphragmatic movement) is preceded by
3A-26s
March 2, 1990
The American Journal of Medicine
Snoring, Hypnotics, and the Elderly
The significance of decreased coordination between diaphragmatic and upper airway muscles during sleep after benzodiazepine intake is poorly documented.
Volume 88 (suppl 3A)
SYMPOSIUM
However, regular snorers (i.e., subjects presenting with only partial airway obstruction) given flurazepam 30 mg at bedtime have OSA during the drug night [15]. In these patients, flurazepam had the same effect as alcohol, i.e., a complete airway obstruction was noted compared with baseline. Considering the body of knowledge currently available on sleeprelated cardiovascular changes associated with OSA and on the interaction between the control of breathing and sedative hypnotics, chronic snoring should be systematically investigated before prescription of a sedative hypnotic, since a partial obstruction might thereby be changed into a complete one. Unfortunately, those most at risk of presenting with OSAthe elderly and, especially? postmenopausal womenare also those who chromeally use these drugs. Elderly subjects, in general, are more likely to present with morbidity; health problems are associated with poor sleep at any age, and this is particularly true when chronic insomnia is found in the elderly. Iatrogenically induced OSA encourages the development of sleep-related cardiac arrhythmias in the elderly. We investigated healthy elderly subjects who had been deprived of sleep on the preceding night or had been given 0.6 mglkg whiskey one hour prior to bedtime. These two conditions were selected because of their depressant effect on control of breathing during sleep. Elderly subjects with a respiratory disturbance index between 5 and 10 (i.e., experiencing five to 10 apneashypopneas per hour of sleep) had a clear increase in the number of sleep-related apneas, and in the case of one patient, salvos of premature ventricular complexes occurred in association with the environmentally induced increased sleep apneas [161. HYPNOTICS AND CORONARY ARTERY DISEASE The direct interaction between sedative hypnotic prescription and cardiovascular functioning during sleep has been little investigated to date. In a study of the effect of flurazepam, triazolam, and temazepam on sleep in middle-aged adults complaining of insomnia but without known cardiac lesions, we have noted no significant cardiovascular problems. However, in clinical practice, sedative hypnotics, including the benzodiazepines, are frequently prescribed for patients with cardiovascular diseases, particularly coronary heart disease (angina, myocardial infarction) and for patients with moderate symptoms of disease who frequently present with a disturbed nocturnal sleep and, at times, sleep-onset insomnia. The influence of benzodiazepines on cardiac function in patients with disturbed sleep can be seen particularly in the development of repetitive apneas during sleep. The occurrence of short apneas would seem to result from purely hemodynamic causes. However, there is some evidence that appears to implicate benzodiazepine hypnotic drugs in their development and continuation. Nine patients treated for myocardial infarction and receiving a benzodiazepine at bedtime during the previous three to six months were monitored polygraphitally during sleep at three different times: at entry, while taking the drug; on the last night of a three-day withdrawal from benzodiazepine; and on the third night after resumption of their prescribed regimen. At baseline with benzodiazepine, mean total sleep
ON INSOMNIA/GUlLLEMlNAULT
time was 342 i 32 minutes, indicating a poor quality of sleep despite hypnotic intake. The mean percentage of REM sleep was 15 +- 3.8 percent. Central apneas of between 10 and 18 seconds’ duration were seen, predominantly during REM sleep, in three patients. The number of central apneas varied between 32 and 151. Two of the three patients presented clusters of short central apneas. During the withdrawal period, patients’ sleep was worse on the third drug-free night: mean total sleep time was 317 2 47 minutes. However, none of the patients studied at this time presented with apneas. When patients resumed benzodiazepine intake, abnormal breathing patterns were again noted in the same three patients as at baseline. The lowest oxygen saturation during the night always appeared in association with the pattern of repetitive central apneas under baseline and drug conditions (mean +- SD were as follows: baseline oxygen saturation, 85 i: 6 percent; off-drug oxygen saturation, 90 +5 percent; on-drug oxygen saturation, 83 t 5.3 percent). The study showed that a breathing pattern, which occasionally led to a brief episode of oxygen desaturation, was associated with benzodiazepine intake, at least in three post-myocardial infarction patients. There is a time lag between the return of oxygenated blood from the right circulation to the left atrium and ventricle (i.e., into the coronary arteries) and that moment when peripheral and central chemoreceptors respond to changes in the blood oxygen content. This delay, short in normal subjects, increases significantly with left ventricular failure and is at least partially responsible for repetitive central apneas or for the development of classic Cheyne-Stokes respiration. However, this mechanism alone does not explain the short, repetitive apneic events seen in some of our patients since the short apneas disappeared on benzodiazepine withdrawal. In persons with existing cardiac lesions, benzodiazepines may also lead to repetitive short central apneas accompanied by brief, more marked, decreases in oxygen saturation. The combination of benzodiazepine and cardiac lesion seems to have an additive effect in certain patients, independent of the mechanisms involved. The effect of these decreases in oxygen saturation on an already compromised heart is still unclear, as is the impact of benzodiazepines on baroreceptors, particularly in populations of at-risk subjects. However, this type of investigation into the impact of benzodiazepine sedation, as well as information on benzodiazepines’ potential impact in patients presenting with cardiac failure, would be helpful to internists and to general practitioners, who face therapeutic dilemmas when treating these cases. COMMENTS Sleep and sleep states are associated with controls of vital functions that differ from those observed during wakefulness. The benzodiazepines, like any psychoactive drugs, have an impact on these controls. Recently, sleep researchers have given attention to the changes induced by CNS-depressant drugs on the control of air exchange during sleep, but little or no information is available on the impact of these drugs on the control of other vital functions during sleep. Nevertheless, epidemiologic studies in the Western
March 2, 1990
The American Journal of Medicine
Volume 88 (suppl 3A)
3A-27s
SYMPOSIUM
ON INSOMNIA I GUILLEMINAULT
Hemisphere indicate that the largest group of longterm consumers of sedative hypnotic drugs is the elderly. Elderly subjects often present with polymorbidity and, undoubtedly, breathing disorders and cardiovascular disease are more common in older age groups; sleep-related worsening of symptoms (due to sleep-state physiology) may lead to disturbed sleep and complaints of insomnia. To avoid biased data (including data that are difficult to interpret), many sleep laboratories study highly selected groups that exclude polymorbidity. These studies are useful; they assess the efficacy of sedative hypnotic drugs m sleep disturbance on welldefined populations of “healthy elderly” presenting only with sleep complaints. Very little investigation, however, has been carried out in populations presenting with ,several health problems, and little has been directed toward the impact of drugs on central controls of vital functions not during wakefulness but during sleep (or night), the time of peak hypnotic drug levels in the blood. COPD patients, cardiac patients, and obese patients assuredly are to be found among the ranks of chronic hypnotic users, who often combine alcohol with sedative intake. In a time of an increasingly aged population, nursing homes are more and more in demand, and families and guardians desire that the elderly sleep well at night without risking a confusional awakening that could lead to abnormal behavior endangering the resident as well as others. Epidemiologic studies also indicate that subjects in nursing homes are in poorer health than other elderly persons and are also long-term users of sedative hypnotic drugs [l]. In a very recent international symposium on sleep and health risks held in Marburg, West Germany, Kripke and Ancoli-Israel [17] reported on a large study performed in San Diego, California, and showed that elderly women who presented with an apnea index above 5 not only had a significantly greater chance of dying at an earlier age than others in their group, but also of dying at night while in bed
3A-28s
March 2, 1990
The American Journal of Medicine
(supposedly asleep). They also received more sedative hypnotic medications. Further investigation of the interaction of sedative hypnotics on the control of vital functions during sleep in health and disease, particularly in the older age groups, is clearly needed. REFERENCES 1. Ancoli-Israel 5. Klauber MR. Krioke DF. Parker L: Increased risk of mottalltv with sleeo apnea in nursing home patients:’ preliminary report (abstr). Sleep Res 1988; 17: 14i. 2. Fleetham J, West P, Mezon B, Conway W, Roth T, Ktyger M: Sleep, arousals, and oxygen desaturation in chronic obstructive pulmonary disease. Am Rev Respir Dis 1982; 126: 429-433. 3. Calverley PMA, Brezinova V, Douglas NJ, et at The effect of oxygenation on sleep quality in chronic bronchitis and emphysema. Am Rev Respir Dis 1982; 126: 206-210. 4. Gaddie J, Legge JS, Palmer KNV, Petrie JC, Wood RA: Effect of nitrazepam in chronic obstructive bronchitis. Br Med J 1972: 2: 688-689. 5. Geddes DM, Rudolf M, Saunders i(B: Effect of nitrazepam and flurazepam on the ventilatory response to carbon dioxide. Thorax 1976; 31: 548-551. 6. Rudolf M, Geddes DM, Turner JAM, Saunders KB: Depression of central respiratory drive bv nitrazeoam. Thorax 1978: 33: 97-100. 7. Cohn MA: Hybnotics and the control of breathing: a review. Br J Ciin Pharmacol 1983; 16 (suppl 2): 245%2505. 8. ClarkTJH, Collins JV, Tong D: Respiratory depression caused by nitrazepam In patients with respiratory failure. Lancet 1971; II: 737-738. 9. Model DG: Nltrazepam induced respiratory depression in chronic obstructive lung disease. Br J Dis Chest 1973; 67: 128-130. 10. Cummiskey J, Guilleminault C, Del Rio G, Silvestri R: The effects of flurazepam on sleep studies in patients with chronic obstructive pulmonary disease. Chest 1983; 84: 143-147. 11. Midgren B, Hansson L, Skeidsvoll H, Elmqvist D: The effect of nitrazepam and flunitrazepam on oxygen desaturation during sleep in patients with stable hypoxemic non-hypercapnic COPD. Chest 1989; 95: 765-768. 12. Guilleminault C, Dement WC (eds): Sleep apnea syndromes. New York: Alan R. Liss, 1978. 13. Ewing DJ: Practical bedslde investigation of diabetic autonomic failure. In: Bannister R, ed. Autonomic failure. London: Oxford University Press, 1983; 371-405. 14. Mancia G, Zanchett A: Cardiovascular regulation during sleep. In: Orem J, Barnes CD, eds. Physiology in sleep. New York Academic Press Inc., 1980; 2-55. 15. Guilleminault C. Cummiskev J. Silvestri R: Benzodiazeoines and resoiration during sleep. In: Udsin E, ‘Clarke P, Tellman D, Greenblatt D, Paul SM, eds. Pharmacology 0; benzodiazepines. London: Macmillan, 1982; 229-236. 16. Guilleminault C, Silvestri R, Mondini S, Coburn S: Aging and sleep apnea: action of benzodiazepine, acetazolamide, alcohol and sleep deprivation in a healthy elderly group. J Gerontol 1984; 39: 655-661. 17. Kripke D, Ancoli-Israel S: Health risk of insomnia (abstr). In: Peter JH, Podzus T, von Wicheri P, eds. Sleep and health risk, an international symposium. Marburg: Phillips Universitat Press, 1989; 90.
Volume 88 (suppl 3A)
