Obesity Research & Clinical Practice (2008) 2, 119—124
ORIGINAL ARTICLE
Weight reduction improves nocturnal respiration in obese sleep apnoea patients—–A randomized controlled pilot study Pia Nerfeldt a,∗, Bengt Y. Nilsson b, Joanna Udd´ en c, Stephan R¨ ossner c, Danielle Friberg a a
Department of Clinical Science, Intervention and Technology, Division of Ear, Nose and Throat Diseases, Karolinska Institute at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden b Department of Clinical Neuroscience, Division of Neurophysiology, Huddinge, Karolinska Institute at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden c Department of Medicine, Huddinge, Division of Endocrinology, Karolinska Institute at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden Received 10 March 2008; accepted 11 March 2008
KEYWORDS Sleep apnoea; Sleep apnoea syndrome; Obstructive sleep apnoea; Sleep disorders; Obesity; Weight loss; Weight reduction; Diet; RCT
Summary Objectives: Randomized controlled pilot study of the effect of weight reduction on nocturnal respiratory parameters in obese patients with obstructive sleep apnoea syndrome (OSAS). Methods: Twenty consecutive obese male patients fulfilling OSAS criteria at Karolinska University Hospital were randomized into two groups. Intervention with an 8-week weight reduction programme consisting of a low-calorie diet, together with group meetings, was evaluated compared to expectancy alone for the control group, followed by a crossover. Follow-up at 3 months included anthropometrics and ambulant sleep apnoea recordings. Results: Eleven of twenty men completed the protocol. There were significant differences between the intervention group (n = 6) and the control group (n = 5) in changes of weight (p < 0.01) and oxygen desaturation index (ODI4 ) (p < 0.05). We also found a significant positive correlation in these 11 males after the crossover between their reduction in weight and their reduction in ODI4 (p < 0.05). Conclusions: This pilot study indicates that weight reduction improves nocturnal respiration in obese OSAS patients after 3 months’ dietary treatment compared to expectancy. © 2008 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
∗ Corresponding author at: Karolinska University Hospital Huddinge, ORL Department B53, SE-141 86 Stockholm, Sweden. Tel.: +46 8 585 80000; fax: +46 8 746 75 51. E-mail address: [email protected] (P. Nerfeldt).
1871-403X/$ — see front matter © 2008 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2008.03.002
120
P. Nerfeldt et al.
Introduction Obstructive sleep apnoea syndrome (OSAS) has a prevalence of about 4% in males and 2% in females [1], and a majority of OSAS patients are overweight [2]. Fat disposition around the pharynx and in the thorax and abdomen is a major cause of OSAS [3,4]. The intermittent obstruction of the pharynx during sleep causes apnoeas and arousals, which result in impaired sleep quality and an often obvious daytime sleepiness [5]. OSAS and obesity are considerable health problems in themselves, and in combination they are a high risk factor for morbidity and mortality, particularly in cardiovascular diseases [6—14]. The treatment of obese OSAS patients is a challenge since their compliance is insufficient and the disease is often lifelong. The effects of weight reduction in obese OSAS patients have been the focus of attention for a long time, and previous studies have shown that a reduction in body weight improves the frequency of apnoea in the short term [15—17]. A reduction in upper airway collapsibility [18] and an increase in the size of the upper airway passage [19] are seen after weight reduction. The use of surgical intervention with different types of gastric banding is becoming more and more common [20], showing good results also on indices of sleep apnoea [19,21—23], though undoubtedly with a considerable risk of complications and, according to some studies, even mortality [24]. Conservative weight reduction involves much less risk as far as morbidity and mortality are concerned, but is not generally considered to be equally successful. The low-calorie diet (LCD) method had not been properly evaluated in this patient group when this study was initiated, although Suratt [25], for example, showed reduced respiratory disturbances after a very low-calorie diet in eight obese subjects in a
non-randomized study as early as 1992. Randomized controlled studies of dietary weight reduction in obese OSAS patients are few and therefore needed [26]. The aim of our study was to evaluate the dietary intervention programme of an LCD in obese OSAS patients, measuring its effect on weight and nocturnal oxygen desaturation index in a randomized control pilot study.
Methods Objective A prospective randomized controlled study of obese untreated OSAS patients, to evaluate the effect of a group programme of LCD weight reduction on weight and respiratory parameters after 3 months.
Subjects and study design Twenty patients fulfilling the inclusion criteria were asked and agreed to participate in the study when seeking medical attention at the Otorhinolaryngological (ORL) Department at Karolinska University Hospital for snoring and daytime sleepiness. They had had ambulant sleep apnoea recordings taken before their visit to the department. Some had failed other OSAS treatment (continuous positive airway pressure (CPAP) or a mandibular retaining device (MRD)) for different reasons. The inclusion criteria were: 20- to 69-year-old males with obesity defined as body mass index (BMI) ≥ 30, apnoea-hypopnoea index (AHI) ≥ 10 and/or oxygen desaturation index (ODI4 ) ≥ 6, together with subjective symptoms of OSAS. Exclu-
Figure 1 Flowchart of the study.
Weight reduction improves nocturnal respiration in OSAS
121
Table 1 Patient characteristics at baseline for the two randomization groups, together with separate data for those completing the study and dropouts Median (range)
Intervention group
Control group
Completing
Dropouts
No. of subjects Age (years) Weight (kg) BMI (kg/m2 )
10 50 (35—69) 120 (100—180) 38 (33—54)
10 48 (28—57) 106 (98—126) 34 (30—36)
11 51 (35—69) 107 (98—126) 36 (30—40)
9 44 (28—55) 116 (104—180) 34 (30—54)
BMI = body mass index.
sion criteria were: serious psychiatric disease, insufficient knowledge of Swedish (preventing participation in group therapy) and other OSAS treatment at the start of the study. The patients were examined and then randomized with sealed envelopes into two groups by an otorhinolaryngologist. A nurse at the obesity unit performed anthropometric data collection. The patients filled in a questionnaire about their symptoms. They then underwent the baseline full-night ambulant sleep apnoea recording. See Fig. 1 for a flowchart of the study. The treatment group started the weight reduction, while the control group was given the advice to maintain their weight. The dietary treatment consisted of an 8-week LCD protein drink (Nutrilett® ) containing approx. 800 kcal daily, together with group meetings once a week supervised by a specialized nurse for support and encouragement. During the first 7 weeks the patients were not allowed to eat anything else, but during the last week they gradually started to eat balanced low-calorie meals. Follow-up visits were performed every second week, including physical check-ups and urine analysis for ketonuria, as a marker of dietary compliance. Following the 8-week LCD, a second sleep apnoea recording was made. Thereafter a crossover took place and the control group was subjected to an LCD programme and group meetings and finally completed a third sleep apnoea recording.
The anthropometrics of the patients are listed in Table 1.
Sleep recordings An ambulatory full-night sleep apnoea recording was made in the patient’s home. It consisted of six channels (oronasal thermistor, mattress showing respiratory and body movements, pulse oximetry, pulse frequency, snoring and body position) (MicroDigitrapper, Synectics Medical, Stockholm, Sweden). The apnoea—hypopnoea index measured by the thermistor was not considered to be a consistently reliable measure of the airflow at this time and was therefore excluded. A blinded qualified neurophysiologist interpreted the recordings.
Definitions of sleep parameters The ODI4 index is the number of events when the pulse oximetry indicates a decline in oxygen saturation by at least 4% points from the patient’s individual baseline divided by the estimated total sleep time, expressed as a number per hour.
Statistical analysis Results are stated as median (range). Nonparametric tests, Mann—Whitney U (MWU), Wilcoxon Signed-Rank (WSR) and Spearman Rank Correlation
Table 2 Significant intra-individual improvement after 3 months’ weight reduction with an LCD for the intervention group in weight, BMI and ODI4 , which was not seen in the control group after expectancy. The last column shows the results after LCD for the control group Intervention group (n = 6) Median kg BMI ODI4
Baseline 115 (100—124) 37 (33—40) 59.5 (8—94)
Control group (n = 5) After LCD
Baseline *
96.5 (90—110) 32.2 (27.8—35.2)* 9.5 (4—71)*
105 (98—126) 33.1 (30.2—36.4) 22 (15—63)
8 week expectancy
After LCD (range)
.
102 (100—127) n.s. 92 (87—114)* 32.6 (30.9—36.7) n.s. 29.4 (26.9—33)* 21 (8—46) n.s. 23 (11—45) n.s.
kg, kilogram (2.2 pounds); BMI, body mass index; ODI4 , oxygen desaturation index ≥ 4%; LCD, low-calorie diet; n.s., non-significant. * Significant reduction from baseline with p < 0.05 WSR test.
122
P. Nerfeldt et al.
(SRC), were used in the statistics. In addition, since the values were normally distributed, a parametric two-sample t-test was added when analysing the difference between the groups in ODI4 improvement.
Ethics The studies were approved by the Regional Ethical Review Board in Stockholm.
Results Eleven patients completed the protocol with intervention and follow-up sleep apnoea recordings. There were significant intra-individual changes in weight and ODI4 in the treatment group (p < 0.05, WSR test), which were not seen in the control group, see Table 2. When comparing the intervention (n = 6) and control (n = 5) groups, a significant interindividual difference was seen in weight reduction (p < 0.01, MWU test). A significant difference was also seen between the groups in ODI4 improvement when using the two-sample t-test (p = 0.045), but only a tendency towards a significant difference in ODI4 improvement (p = 0.067) when using the MWU test, see Fig. 2. After the crossover, there was a significant positive correlation between the reduction in weight and ODI4 in the 11 patients after dietary treatment (p = 0.005, SRC test), see Fig. 3. The median reduction was 11% for weight (p < 0.01) and 28% for ODI4
Figure 2 The changes in ODI4 after a low-calorie diet for the intervention group (LCD diet) and expectancy alone for the control group (expectancy). The differences between the groups are significant at the p = 0.045 level using two-sample t-test and p = 0.067 with the Mann—Whitney U test.
Figure 3 A significant positive correlation between the reduction in weight (dkg) and the reduction in oxygen desaturation index (dODI), p = 0.0053, SRC test.
(p < 0.05). Three out of eleven patients fulfilled the success criteria of an ODI4 reduction of at least 50% to a level
ORIGINAL ARTICLE
Weight reduction improves nocturnal respiration in obese sleep apnoea patients—–A randomized controlled pilot study Pia Nerfeldt a,∗, Bengt Y. Nilsson b, Joanna Udd´ en c, Stephan R¨ ossner c, Danielle Friberg a a
Department of Clinical Science, Intervention and Technology, Division of Ear, Nose and Throat Diseases, Karolinska Institute at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden b Department of Clinical Neuroscience, Division of Neurophysiology, Huddinge, Karolinska Institute at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden c Department of Medicine, Huddinge, Division of Endocrinology, Karolinska Institute at Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden Received 10 March 2008; accepted 11 March 2008
KEYWORDS Sleep apnoea; Sleep apnoea syndrome; Obstructive sleep apnoea; Sleep disorders; Obesity; Weight loss; Weight reduction; Diet; RCT
Summary Objectives: Randomized controlled pilot study of the effect of weight reduction on nocturnal respiratory parameters in obese patients with obstructive sleep apnoea syndrome (OSAS). Methods: Twenty consecutive obese male patients fulfilling OSAS criteria at Karolinska University Hospital were randomized into two groups. Intervention with an 8-week weight reduction programme consisting of a low-calorie diet, together with group meetings, was evaluated compared to expectancy alone for the control group, followed by a crossover. Follow-up at 3 months included anthropometrics and ambulant sleep apnoea recordings. Results: Eleven of twenty men completed the protocol. There were significant differences between the intervention group (n = 6) and the control group (n = 5) in changes of weight (p < 0.01) and oxygen desaturation index (ODI4 ) (p < 0.05). We also found a significant positive correlation in these 11 males after the crossover between their reduction in weight and their reduction in ODI4 (p < 0.05). Conclusions: This pilot study indicates that weight reduction improves nocturnal respiration in obese OSAS patients after 3 months’ dietary treatment compared to expectancy. © 2008 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
∗ Corresponding author at: Karolinska University Hospital Huddinge, ORL Department B53, SE-141 86 Stockholm, Sweden. Tel.: +46 8 585 80000; fax: +46 8 746 75 51. E-mail address: [email protected] (P. Nerfeldt).
1871-403X/$ — see front matter © 2008 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2008.03.002
120
P. Nerfeldt et al.
Introduction Obstructive sleep apnoea syndrome (OSAS) has a prevalence of about 4% in males and 2% in females [1], and a majority of OSAS patients are overweight [2]. Fat disposition around the pharynx and in the thorax and abdomen is a major cause of OSAS [3,4]. The intermittent obstruction of the pharynx during sleep causes apnoeas and arousals, which result in impaired sleep quality and an often obvious daytime sleepiness [5]. OSAS and obesity are considerable health problems in themselves, and in combination they are a high risk factor for morbidity and mortality, particularly in cardiovascular diseases [6—14]. The treatment of obese OSAS patients is a challenge since their compliance is insufficient and the disease is often lifelong. The effects of weight reduction in obese OSAS patients have been the focus of attention for a long time, and previous studies have shown that a reduction in body weight improves the frequency of apnoea in the short term [15—17]. A reduction in upper airway collapsibility [18] and an increase in the size of the upper airway passage [19] are seen after weight reduction. The use of surgical intervention with different types of gastric banding is becoming more and more common [20], showing good results also on indices of sleep apnoea [19,21—23], though undoubtedly with a considerable risk of complications and, according to some studies, even mortality [24]. Conservative weight reduction involves much less risk as far as morbidity and mortality are concerned, but is not generally considered to be equally successful. The low-calorie diet (LCD) method had not been properly evaluated in this patient group when this study was initiated, although Suratt [25], for example, showed reduced respiratory disturbances after a very low-calorie diet in eight obese subjects in a
non-randomized study as early as 1992. Randomized controlled studies of dietary weight reduction in obese OSAS patients are few and therefore needed [26]. The aim of our study was to evaluate the dietary intervention programme of an LCD in obese OSAS patients, measuring its effect on weight and nocturnal oxygen desaturation index in a randomized control pilot study.
Methods Objective A prospective randomized controlled study of obese untreated OSAS patients, to evaluate the effect of a group programme of LCD weight reduction on weight and respiratory parameters after 3 months.
Subjects and study design Twenty patients fulfilling the inclusion criteria were asked and agreed to participate in the study when seeking medical attention at the Otorhinolaryngological (ORL) Department at Karolinska University Hospital for snoring and daytime sleepiness. They had had ambulant sleep apnoea recordings taken before their visit to the department. Some had failed other OSAS treatment (continuous positive airway pressure (CPAP) or a mandibular retaining device (MRD)) for different reasons. The inclusion criteria were: 20- to 69-year-old males with obesity defined as body mass index (BMI) ≥ 30, apnoea-hypopnoea index (AHI) ≥ 10 and/or oxygen desaturation index (ODI4 ) ≥ 6, together with subjective symptoms of OSAS. Exclu-
Figure 1 Flowchart of the study.
Weight reduction improves nocturnal respiration in OSAS
121
Table 1 Patient characteristics at baseline for the two randomization groups, together with separate data for those completing the study and dropouts Median (range)
Intervention group
Control group
Completing
Dropouts
No. of subjects Age (years) Weight (kg) BMI (kg/m2 )
10 50 (35—69) 120 (100—180) 38 (33—54)
10 48 (28—57) 106 (98—126) 34 (30—36)
11 51 (35—69) 107 (98—126) 36 (30—40)
9 44 (28—55) 116 (104—180) 34 (30—54)
BMI = body mass index.
sion criteria were: serious psychiatric disease, insufficient knowledge of Swedish (preventing participation in group therapy) and other OSAS treatment at the start of the study. The patients were examined and then randomized with sealed envelopes into two groups by an otorhinolaryngologist. A nurse at the obesity unit performed anthropometric data collection. The patients filled in a questionnaire about their symptoms. They then underwent the baseline full-night ambulant sleep apnoea recording. See Fig. 1 for a flowchart of the study. The treatment group started the weight reduction, while the control group was given the advice to maintain their weight. The dietary treatment consisted of an 8-week LCD protein drink (Nutrilett® ) containing approx. 800 kcal daily, together with group meetings once a week supervised by a specialized nurse for support and encouragement. During the first 7 weeks the patients were not allowed to eat anything else, but during the last week they gradually started to eat balanced low-calorie meals. Follow-up visits were performed every second week, including physical check-ups and urine analysis for ketonuria, as a marker of dietary compliance. Following the 8-week LCD, a second sleep apnoea recording was made. Thereafter a crossover took place and the control group was subjected to an LCD programme and group meetings and finally completed a third sleep apnoea recording.
The anthropometrics of the patients are listed in Table 1.
Sleep recordings An ambulatory full-night sleep apnoea recording was made in the patient’s home. It consisted of six channels (oronasal thermistor, mattress showing respiratory and body movements, pulse oximetry, pulse frequency, snoring and body position) (MicroDigitrapper, Synectics Medical, Stockholm, Sweden). The apnoea—hypopnoea index measured by the thermistor was not considered to be a consistently reliable measure of the airflow at this time and was therefore excluded. A blinded qualified neurophysiologist interpreted the recordings.
Definitions of sleep parameters The ODI4 index is the number of events when the pulse oximetry indicates a decline in oxygen saturation by at least 4% points from the patient’s individual baseline divided by the estimated total sleep time, expressed as a number per hour.
Statistical analysis Results are stated as median (range). Nonparametric tests, Mann—Whitney U (MWU), Wilcoxon Signed-Rank (WSR) and Spearman Rank Correlation
Table 2 Significant intra-individual improvement after 3 months’ weight reduction with an LCD for the intervention group in weight, BMI and ODI4 , which was not seen in the control group after expectancy. The last column shows the results after LCD for the control group Intervention group (n = 6) Median kg BMI ODI4
Baseline 115 (100—124) 37 (33—40) 59.5 (8—94)
Control group (n = 5) After LCD
Baseline *
96.5 (90—110) 32.2 (27.8—35.2)* 9.5 (4—71)*
105 (98—126) 33.1 (30.2—36.4) 22 (15—63)
8 week expectancy
After LCD (range)
.
102 (100—127) n.s. 92 (87—114)* 32.6 (30.9—36.7) n.s. 29.4 (26.9—33)* 21 (8—46) n.s. 23 (11—45) n.s.
kg, kilogram (2.2 pounds); BMI, body mass index; ODI4 , oxygen desaturation index ≥ 4%; LCD, low-calorie diet; n.s., non-significant. * Significant reduction from baseline with p < 0.05 WSR test.
122
P. Nerfeldt et al.
(SRC), were used in the statistics. In addition, since the values were normally distributed, a parametric two-sample t-test was added when analysing the difference between the groups in ODI4 improvement.
Ethics The studies were approved by the Regional Ethical Review Board in Stockholm.
Results Eleven patients completed the protocol with intervention and follow-up sleep apnoea recordings. There were significant intra-individual changes in weight and ODI4 in the treatment group (p < 0.05, WSR test), which were not seen in the control group, see Table 2. When comparing the intervention (n = 6) and control (n = 5) groups, a significant interindividual difference was seen in weight reduction (p < 0.01, MWU test). A significant difference was also seen between the groups in ODI4 improvement when using the two-sample t-test (p = 0.045), but only a tendency towards a significant difference in ODI4 improvement (p = 0.067) when using the MWU test, see Fig. 2. After the crossover, there was a significant positive correlation between the reduction in weight and ODI4 in the 11 patients after dietary treatment (p = 0.005, SRC test), see Fig. 3. The median reduction was 11% for weight (p < 0.01) and 28% for ODI4
Figure 2 The changes in ODI4 after a low-calorie diet for the intervention group (LCD diet) and expectancy alone for the control group (expectancy). The differences between the groups are significant at the p = 0.045 level using two-sample t-test and p = 0.067 with the Mann—Whitney U test.
Figure 3 A significant positive correlation between the reduction in weight (dkg) and the reduction in oxygen desaturation index (dODI), p = 0.0053, SRC test.
(p < 0.05). Three out of eleven patients fulfilled the success criteria of an ODI4 reduction of at least 50% to a level
