Stephen C. Wilhoit

Effect of very-low-calorie diets with weight loss on obstructive sleep apnea

To determine the effect of very-low-calorie diets (VLCDs) with weight loss on obstructive sleep apnea (OSA), we studied eight obese subjects with OSA, five males and three females. Subjects consumed a VLCD of 1760 kJ (420 kcal) (67% protein, 4% fat, 29% carbohydrate) or 3350 kJ (800 cal) (20% protein, 30% fat, 50% carbohydrate) with 100% of the recommended daily allowance of vitamins and minerals. Mean (+/- SD) values of weight and respiration before and after weight loss were, for weight, 153 +/- 37 and 132 +/- 29 kg (P less than 0.05); for BMI (kg/m2), 54 +/- 13 and 46 +/- 10 (P less than 0.05); for desaturations/h sleep, 106 +/- 50 and 52 +/- 45 (P less than 0.05); for apneas + hypopneas/h sleep, 90 +/- 32 and 62 +/- 49; for Pco2, 48 +/- 10 and 42 +/- 4 torr (P less than 0.05). Desaturation episodes/h and apnea + hypopneas/h improved in six patients. The most obese subject (female, BMI 81) who lost the most weight (47 kg) did not improve, nor did the subject who lost the least weight, 7 kg. The number of movements + arousals from sleep decreased in all patients (P less than 0.05). We conclude that VLCD with weight loss can produce improvement in OSA; subjects who lose a small amount of weight or subjects who are extraordinarily obese before and after weight loss may not improve.

Comparison of Indices Used to Detect Hypoventilation during Sleep

Because there is no uniform method of measuring the severity of sleep apnea, we compared respiratory indices calculated from airflow and oxyhemoglobin saturation (SaO2) signals in 16 subjects during a nights sleep. Airflow was measured with a loosely fitting pneumotachograph or thermister and the following indices calculated manually: total apneas (A) and hypopneas (H); A and H per hour of sleep; total A H time; average duration of A H episodes, and A H index (the product of 2 and 4). SaO2 was measured with a Hewlett-Packard ear oximeter and the following indices calculated with a microcomputer; total number of desaturation (D) episodes; D episodes per hour of sleep; average maximum D; D index (the product of 7 and 8); SaO2 50 (1), and SaO2 10(1). There was a significant correlation among all indices; the highest correlation was between total A and H and total D (r = 0.97). We conclude that SaO2 indices calculated with a microcomputer correlate well with flow indices.

A microcomputer system for monitoring and analysing oxyhemoglobin saturation during sleep.

A computerized data acquisition and analysis routine was developed to quantitate respiratory disturbances in sleeping patients. Polysomnographic recordings of patients consisted of electroencephalograms, electro-occulograms, submental electromyograms, air flow at the nose and mouth, esophageal pressure, and oxyhemoglobin saturation (SaO2). SaO2, a physiological effect of ventilatory airflow, was sampled every two seconds and stored on disk during the nights study for subsequent analysis. Wild points in the data file can be marked so that they will be skipped during analysis. Patient polysomnographs were scored manually for sleep stage by a sleep technician. A file was then created containing the scored sleep information with time marks corresponding to each change in sleep stage during the study. An analysis routine used this file to develop indices of sleep apnea, severity for combinations or specific stages of sleep. These indices were: (A) number of oxyhemoglobin desaturation episodes per hour; (B) average maximum desaturation per episode; and (C) desaturation index, the product of (A) and (B). A graph was plotted showing cumulative sleep time at given SaO2 values. The degree of sleep apnea can be determined using these indices.

Treatment of obstructive sleep apnea with continuous nasal airflow.

Although nasal continuous positive airway pressure (CPAP) is effective therapy for obstructive sleep apnea (OSA), it requires a customfitted nasal appliance and large cumbersome tubing. We therefore designed and tested a new device (NFLOW) to deliver airflow to the nose of patients with OSA. We studied 13 patients the first night without treatment and the following night with NFLOW. The degree of sleep apnea was assessed by the number of desaturations per hour of sleep and the average maximum desaturation per episode. Treatment with NFLOW significantly decreased all parameters (P<0.01) in 9 patients (69%) who tolerated treatment flow rates above 30 LPM. REM sleep time significantly increased with NFLOW use, however, other sleep stage times were not significantly altered. Obstructive apneas ceased in all but 2 of these patients with treatment. Four patients did not tolerate flow rates above 35 LPM and did not improve. We conclude that NFLOW treatment significantly decreases the number of oxyhemoglobin desaturations and improves oxyhemoglobin saturation in patients with OSA who tolerate the procedure.