Long-Term Neurophysiologic Impact of Childhood Sleep Disordered Breathing on Neurocognitive Performance

Stuart F. Quan, M.D.^ab

Stuart_Quan@hms.harvard.edu

Kristen Archbold, Ph.D.^c

Karchbold@nursing.arizona.edu

Alan S. Gevins, D.Sc.^d

Alan@eeg.com

James L. Goodwin, Ph.D.^a

jamieg@arc.arizona.edu

aArizona Respiratory Center, University of Arizona College of Medicine, Tucson, AZ, bDivision of Sleep Medicine, Harvard Medical School, Boston, MA, cPractice Division, University of Arizona College of Nursing, Tucson, AZ, dSAM Technology & San Francisco Brain Research Institute, San Francisco, CA

Abstract

Study Objective. To determine the impact of sleep disordered breathing (SDB) in children on neurocognitive function 5 years later.

Design, Setting, and Participants. A subgroup of 43 children from the Tucson Children’s Assessment of Sleep Apnea Study (TuCASA) who had SDB (RDI > 6 events/hour) at their initial exam (ages 6-11 years) were matched on the basis of age (within 1 year), gender and ethnicity (Anglo/Hispanic) to 43 children without SDB (Control, RDI < 4 events/hour). The Sustained Working Memory Task (SWMT) which combines tests of working memory (1-Back Task), reaction time (Simple Reaction Time) and attention (Multiplexing Task) with concurrent electroencephalographic monitoring was administered approximately 5 years later.

Results. There were no differences in performance on the working memory, reaction time and attention tests between the SDB and Control groups. However, the SDB group exhibited lower P300 evoked potential amplitudes during the Simple Reaction Time and Multiplexing Tasks. Additionally, peak alpha power during the Multiplexing Task was lower in the SDB Group with a similar trend in the Simple Reaction Time Task (p=0.08).

Conclusions. SDB in children may cause subtle long-term changes in executive function that are not detectable with conventional neurocognitive testing and are only evident during neuroelectrophysiologic monitoring.

Introduction

There is increasing evidence that childhood sleep disordered breathing (SDB) is associated with neurobehavioral morbidity (1-3). In cross-sectional studies, children with SDB are found to have deficits in a variety of neurobehavioral domains including attention, executive function, behavior regulation, alertness, learning and academic performance (3). Treatment of OSA with either tonsillectomy or adenoidectomy often results in resolution or improvement in many of these domains (4, 5).

Despite the large amount of data implicating SDB as a causative factor in producing deficits in neurocognition in children, there have been few studies implicating SDB in children as a risk for long-term neurobehavioral morbidity. Several studies have reported that snoring as a surrogate for SDB predicted increased risk for hyperactive behavior (6-8). In addition, in a retrospective analysis, Gozal and Pope (9) reported that low performing middle school students had a greater likelihood of snoring during childhood than their high performing classmates. However, there have been no studies of long-term neurobehavioral morbidity that have used polysomnography (PSG) to document the presence of SDB. Determining whether long-term or permanent deficits in neurocognition occur as a result of SDB will be important in timing of treatment intervention in these children.

In the present study, a subset of the Tucson Children's Assessment of Sleep Apnea Study (TuCASA) underwent additional cognitive neurophysiological testing to determine whether SDB documented during childhood was a risk factor for deficits 5 years later. We hypothesized that children with SDB would exhibit subtle abnormalities during these neurophysiologic tests.

Materials & Methods

Subjects. The Tucson Children’s Assessment of Sleep Apnea study (TuCASA) was a longitudinal cohort established to investigate the correlates and natural history of childhood sleep disordered breathing. Recruitment and overall study methods have been previously described (10, 11). In brief, the TuCASA cohort consisted of healthy school-aged children that were enrolled in a large urban school district in the Southwest United States. With the cooperation of their respective elementary schools, parents of the students were asked to complete a brief screening questionnaire and to provide contact information if they wanted to see if their child was eligible for the study. Those who qualified were then studied using a single overnight unattended in-home PSG along with completion of a questionnaire regarding their sleep habits. TuCASA initially recruited 503 participants (ages 6 – 11) who had their PSGs recorded between the years of 2000 and 2004. Approximately 5 years later, the study attempted to contact the same participants and was successful in restudying 319 children who had valid PSGs for both the baseline and follow-up time points. From this group, we selected 2 groups of children based on their respiratory disturbance index measured from the PSG performed during their baseline exam cycle. No children had received continuous positive airway pressure treatment for SDB. The first group (SDB) consisted of children with SDB as defined by a respiratory disturbance index (RDI) > 6 /hour. The second group (Control) was children without SDB as defined by a RDI < 4/hour. Each child in the Control group was matched to a child in the SDB group on the basis of age (within 1 year), gender and ethnicity. This resulted in a study cohort of 43 pairs of children.

All methods used to recruit subjects and to collect the present data set were approved both by the University of Arizona Human Subjects Committee and the Tucson Unified School District Research Committee. In all cases, we obtained written informed consent from the parents, and assent from the children.

Study Design. Children from both groups were asked to return to the TuCASA sleep laboratory to undergo the Sustained Working Memory Test (SWMT) which was adapted for use in children. The SWMT is an automated cognitive neurophysiological test that combines cognitive test performance measures with electroencephalograhic (EEG) measures. It has been validated in distinguishing cognitive performance in subjects who have ingested alcohol, caffeine, diphenhydramine and who have been sleep deprived (12, 13). These studies were performed on a day separate from any other testing done for TuCASA.

Polysomnography. In both the baseline and follow-up examinations, children underwent unattended, nocturnal home PSG using the Compumedics PS-2 system (Abbotsford, Victoria, Australia) (10, 11). The following signals were obtained: C3/A2 and C4/A1 EEG, right and left electrooculogram, a bipolar submental electromyogram, thoracic and abdominal displacement (inductive plethysmography bands), airflow (nasal/oral thermocouple), nasal pressure, electrocardiogram (single bipolar lead), snoring (microphone attached to a vest), body position (Hg gauge sensor), pulse oximetry (Nonin, Plymouth, MN) and ambient light (sensor attached to the vest to record on/off). Using Compumedics W-Series Replay, v 2.0, release 22, sleep stages were scored according to Rechtschaffen and Kales criteria (14). The RDI was defined as the number of respiratory events (apneas and hypopneas) per hour of the total sleep time irrespective of any associated oxygen desaturation or arousal. Studies with less than 4 hours of scorable oximetry were classified as failed studies and were repeated if the participant consented. Central apneas were scored if both airflow and thoracoabdominal effort were absent. However, central events that occurred after movement were not included. Obstructive apneas were identified if the airflow signal decreased to below 25% of the “baseline amplitude”. Hypopneas were scored if the magnitude of any ventilation signal decreased below approximately 70% of the “baseline” amplitude, as described previously (15).

Body Mass Index Computation. Height and weight were collected on a platform scale. BMI was calculated kg/m², and percentile of BMI adjusted for age, sex and ethnicity was calculated with a standardized data-analysis program from the Centers for Disease Control (http://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm ).

Wechsler Abbreviated Scale of Intelligence (WASI). The WASI (16) is nationally standardized intelligence test which is linked to the Wechsler Intelligence Scale for Children®—Fourth Edition (WISC–IV®). It was administered in TuCASA as part of an overall neurocognitive test battery within several weeks of the SWMT.

Sustained Working Memory Test. The SWMT (17) consists of a brief 25 minute computerized test consisting of two blocks of an attentional multiplexing task, an easy and a more difficult version of a spatial n-back working memory task, and eyes open and eyes closed resting tasks. The test is designed for concurrent EEG recording. Data collected included EEG and evoked potential (EP) signals, as well as task performance measures. All subjects were trained how to perform each task the same day the test was administered.

In the attentional multiplexing task (MT), the participants were required to monitor multiple stimuli as they changed shape, color, and pattern, and to sort each object into a bin based on its relevant features. This task adapted to an individual’s ability level; task difficulty increases if performance exceeds a pre-defined threshold, and decreases if performance falls below the threshold. Each MT block lasted approximately 3.5 minutes. The working memory (WM) test consisted of a 3.5 minute spatial 1-back task, in which participants compared the location of the dot stimulus on each trial to that on the immediately preceding trial. A simple reaction time (SRT) test with the same stimulus and response characteristics also was administered as a control task. Resting EEG was also recorded for 1.5 min with eyes open and 1.5 min with eyes closed.

EEGs were recorded from seven scalp locations (Fz, F3, F4, Cz, Pz, P3, P4) positioned via a nylon electrode cap and referenced to linked mastoids. This montage was designed to include adequate spatial representation of the signal features of primary interest as defined by prior high-resolution EEG studies of the working memory tasks used herein (18). Potentials generated by eye movements and blinks were recorded by electrodes positioned above and at the outer canthus and superior orbital ridge of each eye. The resulting data were digitally high-pass filtered at 0.5 Hz. EEG was recorded continuously during task performance and during passive resting conditions. Electrode impedances were kept < 5 KΩ for the references and <20 KΩ on all other channels.

Automated artifact-detection and artifact-decontamination filters were used to minimize contaminants induced by eye movement and other physiologic and instrumental sources. All data were then visually inspected, and any residual contaminants were excluded from further analysis.

SWMT Data Analysis. To assess neurophysiological measures between the SDB and Control groups, EEG power spectra and EPs were calculated. Power spectra were computed from all artifact-free EEG for each task block and converted to dB power with a log10 transformation. To calculate EPs, trials were averaged in 1.2 s epochs beginning 0.2 s before stimulus onset. EP peak amplitudes were measured relative to the mean amplitude in the prestimulus interval.

A number of prior studies have served to identify spectral features of the EEG that are sensitive to task-difficulty manipulations in the types of working memory tasks used in the SWMT Exam (18-20). Based on such previous findings, a number of such sensitive EEG and EP signals were compared between the SDB and Control groups. EEG alpha power was measured as the maximum power in a 2 Hz band between 8- to 12-Hz in all tasks. Amplitude of the P300 (measured in a 100 ms window centered on the largest positive peak between 250 and 520 ms at Pz), and slow-wave EPs (measured in a 250 ms window centered on the largest positive peak between 250 and 650 ms) were computed in the SRT and 1-back WM tasks. Because of the nature of the MT task, EPs were calculated relative to the onset of the visual feedback that immediately follows a correct or incorrect response. A P300 was measured in the MT as the largest positive peak occurring 200-450 ms after the feedback.

Statistical Methods. Potential differences in the characteristics of the study population between the SDB and Control groups were evaluated using Students’ unpaired t-test or linear correlation. Inasmuch as intelligence is a significant factor in determining performance on neurocognitive tests, analysis of covariance was used to compare performance on the various components of the SWMT between groups while controlling for intelligence as assessed by the WASI. Other covariates were included in the models if significant on univariate analyses. Data were analyzed with IBM SPSS Version 20 (http://www-01.ibm.com/software/analytics/spss/) and are presented as mean + SE.

Results

In Table 1 is shown the characteristics of the children in this study.

Boys and Anglos comprised the majority of the study cohort. No children had undergone an adenotonsillectomy at the time of the 1^st PSG, and only 2 had this procedure during the time interval before the 2^nd PSG. As defined by the study design, there were no differences between the SDB and Control groups with respect to age, the time of the 1st or 2nd PSG, or at time of SWMT. Additionally, as dictated by the study design, RDI at the 1st PSG was significantly greater in the SDB group as was the BMI and the standardized BMI (sBMI). The RDI at the 2nd PSG also was higher in the SDB group. However, RDI decreased in both groups over the time period from the 1st to the 2nd PSG. Seven children had SDB on both PSGs and 2 children developed SDB over the study interval. In the SDB group, the mean RDI at the 2nd PSG was below the RDI threshold used to define the Control group at the baseline examination (1st PSG). Significant, but weak negative correlations were observed between sBMI and some of the EEG and evoked potential components of the SWMT [1 back peak alpha, r=-.23, p=0.03; Multiplex Block 1 Alpha Power, r=-.25, p=0.02; Multiplex Block 2 Alpha Power, r=-.21, p=0.05; eyes closed peak alpha, r=-.29, p<0.01; eyes open delta theta power, r=-.25, p=0.02]. Overall, the WASI indicated that the cohort was above average in intelligence and there were no differences between the 2 groups. However, there was considerable heterogeneity within the overall cohort (Minimum WASI: 77; Maximum WASI: 138).

The results of various components of the SWMT are shown in Table 2.

Increased slow eye movement, increased delta/theta band power, and decreased eye closed to eyes open alpha power ratio are neurophysiological indicators of decreased alertness. No differences between the SDB and Control groups were observed for any of these alertness measures.  Similarly, there were no differences with respect to either % items correct or reaction time for the Simple Reaction Time, 1 Back or Multiplexing Tasks. However, the 1 back slow wave amplitude was lower in the SDB Group, and there was a strong trend for the P300 evoked potential amplitude during the Multiplexing Task (p=0.06) and peak alpha power during the Simple Reaction Time Task (p=0.08) to be lower as well. In addition, peak alpha power during the both blocks of the Multiplexing Task was lower in the SDB Group.

Additional analyses were performed to determine whether performance on the SWMT was related to the presence of SDB at the time of the 2^nd PSG. No differences were observed between children who had SDB on the 2^nd PSG and those who did not.

Discussion

In this study, we have demonstrated that after approximately 5 years, several conventional measures of executive function were not different between children with and without SDB. However, neuroelectrophysiologic assessments recorded during task performance were able to distinguish between these 2 groups. These data suggest that SDB in children can have a long-term, albeit subtle impact on neurocognition in children.

Two important domains of executive function are attention and working memory. In our study these were assessed using a simple reaction time task, a multiplexing task and a 1 back working memory task. Although we did not observe that children with SDB had worse performance in either of these domains, previous cross-sectional studies in children have found deficits using a variety of instruments (3). However, many of these studies assessed children derived from clinic populations. In addition, none determined if there was any impact on long-term performance.

The principal finding from our study is that peak alpha power during the multiplexing and simple reaction time tasks was lower in the SDB group. Alpha power reduction is generally considered a marker of cortical activation. Thus, during task performance, peak alpha power should decline as a function of the amount of effort needed to accomplish a given task (21). It is possible that SDB children in this study may have expended more effort to maintain task performance, as evidenced by a lower alpha power. Using the SWMT, similar findings have been observed after marijuana smoking (22).

Similar to the differences in peak alpha power between our SDB and Control groups, we also observed that the P300 evoked potential amplitude during the multiplexing task and the slow wave evoked potential amplitudes during the 1 back task were lower in the SDB group. These evoked potential components are thought to represent aspects of memory encoding, manipulation and retrieval (23). Thus, these data suggest that children with SDB may experience subtle long-term impairment in memory function.

There are several possible explanations of why we did not observe any overt deficits in executive function in children with SDB. First, there was a significant improvement in the RDI in the approximately 5 year interval between the 1st PSG and the testing of these children. Thus, in many of the children, remission of their SDB occurred leading to a reduction in any possible impact of SDB on neurocognition. This would support the contention that overt neurocognitive deficits produced by SDB in school-aged children resolve if SDB improves. Second, the cohort overall had above average intelligence. It is plausible that any impact of SDB would be more evident in those who have less cognitive abilities. Third, it is possible that the 1-back working memory task used in the study was not sufficiently difficult to expose any underlying impairments in executive function. Finally, there is the possibility that inherent cognitive reserve is mitigating the impact of SDB. The cognitive reserve theory postulates that individual differences in how the brain processes tasks may prevent greater insult by using preexisting cognitive processes or by recruiting compensatory ones before there is a detrimental impact on performance (24). Inasmuch as children have the potential for a high amount of neural plasticity, this explanation may be highly relevant in children with SDB.

Our study is not without some important limitations. First, classification of these children into Control and SDB groups was done without regards to desaturation during apnea and hypopnea events. Thus, the impact of oxygen desaturation cannot be assessed. It is unlikely, however, that this is a major confounder because significant oxygen desaturation below 90% was uncommonly observed in these children. Second, there may have been misclassification of children into the SDB and Control groups especially at the cutpoint boundary. We believe this is less likely because by setting the Control cutpoint at < 4 events per hour and the SDB cutpoint at > 6 events per hour, there would have been less risk of misclassification. Third, we did not use intelligence as a factor in assigning children to the 2 groups. However, we believe this had little impact on our results because there was no difference between the groups on the WASI, and we controlled for intelligence in the analyses. Finally, we also observed that body mass index was negatively correlated with performance on some of the SWMT components. Others have found that neurocognitive performance may be negatively associated with obesity (25). However, most studies have been cross-sectional and thus the directionality and causal mechanisms of this association are unclear. Nevertheless, it is unlikely that our results can be explained by this association inasmuch as we controlled for BMI in our analyses. Despite these aforementioned limitations, our study is the only one to our knowledge that has simultaneously assessed neuroelectrophysiologic function during performance of executive function testing in children thus demonstrating its feasibility and potential for acquiring unique information.

In conclusion, SDB in children has the potential to result in subtle long-term detrimental effects in executive function that are not detectable with conventional neurocognitive testing, but may be evident with simultaneous neuroelectrophysiologic monitoring. These data emphasize the importance of recognizing and treating SDB in children in order to prevent possible long-term consequences in neurocognitive function.

Acknowledgements

TuCASA was supported by HL62373. In addition, development of the SWMT was supported by grants from the National Institute of Neurological Diseases and Stroke, The National Institute of Mental Health, The Air Force Research Laboratory and The Office of Naval Research.

Conflicts of Interests: The authors do not have any conflicts of interest to disclose.

References

1. Owens JA. Neurocognitive and behavioral impact of sleep disordered breathing in children. Pediatr Pulmonol. 2009;44(5):417-422.  [CrossRef] [PubMed]
2. Grigg-Damberger M, Ralls F. Cognitive dysfunction and obstructive sleep apnea: from cradle to tomb. Curr Opin Pulm Med. 2012;18(6):580-587.  [CrossRef] [PubMed]
3. Beebe DW. Neurobehavioral morbidity associated with disordered breathing during sleep in children: a comprehensive review. Sleep. 2006;29(9):1115-1134. [PubMed]
4. Giordani B, Hodges EK, Guire KE, et al. Changes in neuropsychological and behavioral functioning in children with and without obstructive sleep apnea following Tonsillectomy. J Int Neuropsychol Soc. 2012;18(2):212-222.  [CrossRef] [PubMed]
5. Ikeda FH, Horta PA, Bruscato WL, Dolci JE. Intellectual and school performance evaluation of children submitted to tonsillectomy and adenotonsillectomy before and after surgery. Braz J Otorhinolaryngol. 2012;78(4):17-23. [PubMed]
6. Chervin RD, Ruzicka DL, Archbold KH, Dillon JE. Snoring predicts hyperactivity four years later. Sleep. 2005;28(7):885-890. [PubMed]
7. Ali NJ, Pitson D, Stradling JR. Natural history of snoring and related behaviour problems between the ages of 4 and 7 years. Arch Dis Child. 1994;71(1):74-76. [PubMed]
8. Urschitz MS, Eitner S, Guenther A, et al. Habitual snoring, intermittent hypoxia, and impaired behavior in primary school children. Pediatrics. 2004;114(4):1041-1048.  [CrossRef] [PubMed]
9. Gozal D, Pope DW,Jr. Snoring during early childhood and academic performance at ages thirteen to fourteen years. Pediatrics. 2001;107(6):1394-1399. [CrossRef]  [PubMed]
10. Goodwin JL, Enright PL, Kaemingk KL, et al. Feasibility of using unattended polysomnography in children for research--report of the Tucson Children's Assessment of Sleep Apnea study (TuCASA). Sleep. 2001;24(8):937-944. [PubMed]
11. Goodwin JL, Vasquez MM, Silva GE, Quan SF. Incidence and Remission of Sleep-Disordered Breathing and Related Symptoms in 6- to 17-Year Old Children-The Tucson Children's Assessment of Sleep Apnea Study. J Pediatr. 2010;  [CrossRef] [PubMed]
12. Gevins A, McEvoy LK, Smith ME, et al. Long-term and within-day variability of working memory performance and EEG in individuals. Clin Neurophysiol. 2012;123(7):1291-1299.  [CrossRef] [PubMed]
13. Smith ME, McEvoy LK, Gevins A. The impact of moderate sleep loss on neurophysiologic signals during working-memory task performance. Sleep. 2002;25(7):784-794. [PubMed]
14. Rechtschaffen A, Kales A. A Manual of Standardized Terminology,
15. Techniques and Scoring System for Sleep Stages of Human Subject. Washington, D.C.: US Government Printing Office, National Institute of Health Publication, 1968
16. Kaemingk KL, Pasvogel AE, Goodwin JL, et al. Learning in children and sleep disordered breathing: findings of the Tucson Children's Assessment of Sleep Apnea (tuCASA) prospective cohort study. J Int Neuropsychol Soc. 2003;9(7):1016-1026.  [CrossRef] [PubMed]
17. Wechsler D. Wechsler Abbreviated Scale of Intelligence. San Antonio, TX: Psychological Corp., 1999
18. Gevins A, Smith ME, McEvoy LK, et al. A cognitive and neurophysiological test of change from an individual's baseline. Clin Neurophysiol. 2011;122(1):114-120.  [CrossRef] [PubMed]
19. Gevins A, Smith ME, McEvoy L, Yu D. High-resolution EEG mapping of cortical activation related to working memory: effects of task difficulty, type of processing, and practice. Cereb Cortex. 1997;7(4):374-385. [CrossRef] [PubMed]
20. Gevins A, Cutillo B. Spatiotemporal dynamics of component processes in human working memory. Electroencephalogr Clin Neurophysiol. 1993;87(3):128-143. [CrossRef] [PubMed]
21. McEvoy LK, Smith ME, Gevins A. Dynamic cortical networks of verbal and spatial working memory: effects of memory load and task practice. Cereb Cortex. 1998;8(7):563-574. [CrossRef] [PubMed]
22. Gevins A, Smith ME, McEvoy LK. Tracking the cognitive pharmacodynamics of psychoactive substances with combinations of behavioral and neurophysiological measures. Neuropsychopharmacology. 2002;26(1):27-39. [PubMed] [PubMed]
23. Hart CL, Ilan AB, Gevins A, et al. Neurophysiological and cognitive effects of smoked marijuana in frequent users. Pharmacol Biochem Behav. 2010;96(3):333-341.  [CrossRef] [PubMed]
24. Polich J. Updating P300: an integrative theory of P3a and P3b. Clin Neurophysiol. 2007;118(10):2128-2148.  [CrossRef] [PubMed]
25. Stern Y. Cognitive reserve. Neuropsychologia. 2009;47(10):2015-2028.  [CrossRef] [PubMed]
26. Liang J, Matheson BE, Kaye WH, Boutelle KN. Neurocognitive correlates of obesity and obesity-related behaviors in children and adolescents. Int J Obes (Lond). 2013 Aug 5.  [Epub ahead of print] [CrossRef] [PubMed]

Reference as: Quan SF, Archbold K, Gevins AS, Goodwin JL. Long-term neurophysiologic impact of childhood sleep disordered breathing on neurocognitive performance. Southwest J Pulm Crit Care. 2013;7(3):165-75. doi: http://dx.doi.org/10.13175/swjpcc110-13 PDF

The Impact of Sleep-Disordered Breathing on Body Mass Index (BMI): The Sleep Heart Health Study (SHHS)

Mark A. Brown, M.D. ¹

James L. Goodwin, Ph.D.²

Graciela E. Silva, Ph.D, MPH.³

Ajay Behari, M.D.⁴

Anne B. Newman, M.D., M.P.H^5,6

Naresh M. Punjabi, M.D., Ph.D.⁷

Helaine E. Resnick, Ph.D., M.P.H.⁸

John A. Robbins, M.D., M.S.H.⁹

Stuart F. Quan, M.D.^2,10

¹Department of Psychiatry, Kaiser Permanente, Portland, OR (markbrownmd@gmail.com); 

²Sleep and Arizona Respiratory Centers, University of Arizona College of Medicine, Tucson, AZ(jamieg@arc.arizona.edu);

³College of Nursing & Health Innovation, Arizona State University, Tempe, AZ (Graciela.Silva@asu.edu); 

⁴Pulmonary and Critical Care Associates of Baltimore, Baltimore, MD (ajaybehari@yahoo.com);

⁵Graduate School of Public Health, Department of Epidemiology, University of Pittsburgh, Pittsburgh, PA

⁶Division of Geriatric Medicine, University of Pittsburgh School of Medicine, Pittsburgh, PA (NewmanA@edc.pitt.edu);

⁷Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, MD (npunjabi@jhmi.edu);

⁸American Association of Homes and Services for the Aging, Washington, DC (heresnick@gmail.com); 

⁹Center for HealthCare Policy and Research, University of California, Davis, Sacramento, CA (jarobbins@ucdavis.edu);

¹⁰Division of Sleep Medicine, Harvard Medical School, Boston, MA (squan@arc.arizona.edu)

Address for correspondence and reprint requests: Stuart F. Quan, M.D., Division of Sleep Medicine, Harvard Medical School, 401 Park Dr., 2^nd Floor East, Boston, MA 02215, Tel (617) 998-8842, Fax (617) 998-8823, Email: squan@arc.arizona.edu

Conflict of Interest Statement: None of the authors have conflicts of interest pertinent to the subject matter of this manuscript. 

Reference as: Brown MA, Goodwin JL, Silva GE, Behari A, Newman AB, Punjabi NM, Resnick HE, Robbins JA, Quan SF. The impact of sleep-disordered breathing on body mass index (BMI): the sleep heart health study (SHHS). Southwest J Pulm Crit Care 2011;3:159-68. (Click here for PDF version of the manuscript)

Abstract

Introduction: It is well known that obesity is a risk factor for sleep-disordered breathing (SDB). However, whether SDB predicts increase in BMI is not well defined. Data from the Sleep Heart Health Study (SHHS) were analyzed to determine whether SDB predicts longitudinal increase in BMI, adjusted for confounding factors.

Methods: A full-montage unattended home polysomnogram (PSG) and body anthropometric measurements were obtained approximately five years apart in 3001 participants. Apnea-hypopnea index (AHI) was categorized using clinical thresholds: < 5 (normal), ≥ 5 to <15 (mild sleep apnea), and ³ 15 (moderate to severe sleep apnea). Linear regression was used to examine the association between the three AHI groups and increased BMI. The model included age, gender, race, baseline BMI, and change in AHI as covariates.

Results: Mean (SD) age was 62.2 years (10.14), 55.2% were female and 76.1% were Caucasian. Five-year increase in BMI was modest with a mean (SD) change of 0.53 (2.62) kg/m² (p=0.071). A multivariate regression model showed that subjects with a baseline AHI between 5-15 had a mean increase in BMI of 0.22 kg/m² (p=0.055) and those with baseline AHI ≥ 15 had a BMI increase of 0.51 kg/m² (p<0.001) compared to those with baseline AHI of <5.

Conclusion: Our findings suggest that there is a positive association between severity of SDB and subsequent increased BMI over approximately 5 years. This observation may help explain why persons with SDB have difficulty losing weight.

Key Words: Sleep Apnea, Weight Gain, Obesity

Abbreviation List: PSG-polysomnogram, SDB-sleep disordered breathing, AHI-apnea hypopnea index, SHHS-Sleep Heart Health Study, BMI-body mass index, SD-standard deviation, SEM-standard error of the mean, ANOVA-analysis of variance

Introduction

There is overwhelming epidemiological and clinical data indicating that obesity is a risk factor for sleep disordered breathing (SDB).^1-8 The association between obesity and SDB is substantial, with high body mass index (BMI) contributing to moderate to severe SDB in 58% of affected persons.⁹ The effect of obesity is greater in men than women^1,10-12 although it decreases with increasing age.^6,7 In addition, weight loss has been demonstrated to decrease the severity of SDB.^10,13,14 Longitudinal data from population studies including the Sleep Heart Health Study (SHHS),¹⁰ the Wisconsin Sleep Cohort,¹⁵ and the Cleveland Family Study⁶ have initially focused on the impact of increased weight on SDB severity. However, examination of the opposite causal pathway has yet to be prospectively addressed.

Anecdotally, patients with SDB appear to have more difficulty losing weight than obese patients without SDB. They also report marked weight gain prior to confirmation of their diagnosis. Two small studies support these empiric observations.^5,16  Given this limited information on the impact of SDB on BMI, data from SHHS was analyzed to examine the impact of SDB on BMI after controlling for change in AHI and severity of SDB.

Methods

Study Design and Population. The SHHS is a multi-center, community-based prospective cohort study of the natural history and cardiovascular consequences of SDB. Details of the study design, sampling, and procedures have been reported.¹⁷ Briefly, between November 1995 and January 1998 participants were recruited from several ongoing prospective cohort studies--the Framingham Offspring and Omni Studies, the Atherosclerosis Risk in Communities Study, the Cardiovascular Health Study, the Strong Heart Study, and the cohort studies of respiratory disease in Tucson and of hypertension in New York. Participants were eligible if they were ≥ 40 years of age and were not being treated for sleep apnea with positive pressure therapy, an oral appliance, oxygen, or a tracheostomy. Habitual snorers < 65 years were over sampled to increase the prevalence of obstructive sleep apnea. Subjects were required to provide written consent and the protocol was approved by the institutional review boards of each of the eight investigative sites. 

Data Collection. A total of 6,441 subjects completed the baseline polysomnogram (PSG), and 4,586 consented to have a second evaluation approximately five years later. This analysis focuses on the 3,040 participants who had PSG and BMI data at both time points. Data from all 215 participants who had a follow-up PSG from the New York center were excluded because they did not meet quality standards for the follow-up examination. The remaining participants died, were too ill to participate, refused to participate, were lost to follow-up or had incomplete covariate data such as weight. This latter group had a higher percentage of Whites (85%) compared with the study group (75.5%) (p-value <0.001). There also were statistically significant differences in baseline BMI, baseline AHI, and age between the study group compared with the excluded group, however, these differences were very small and were not clinically significant. There was no gender difference between the two groups.

Weight was measured on the night of the PSG examination with the participant in light clothes on a calibrated portable scale. Height was obtained at the baseline home visit if not already measured within + 3 months of the parent study. BMI was calculated as weight in kilograms divided by the square of height in meters. Baseline height was used for baseline and follow-up BMI calculations. Age, sex, and ethnicity were self-reported. 

The PSG was conducted using a portable monitor (PS-2 System; Compumedics Limited, Abbotsford, Victoria, Australia), using methods previously described.¹⁸ Apnea was present if there was an absence or near absence of airflow or thoracoabdominal movement (at least < 25% of baseline) for > 10 seconds. Hypopnea was defined as a decrease in the amplitude of the airflow or thoracoabdominal movement below 70% of baseline for > 10 seconds. The apnea-hypopnea index (AHI) was calculated as the number of apnea and hypopnea events, each associated with at least a 4% decrease in oxygen saturation, divided by total sleep time in hours.

Results

Participant characteristics are provided in Table 1. 

Table 1: Characteristics of participants of the Sleep Heart Health Study cohort with complete baseline and follow-up polysomnography and weight measurements as a function of sleep apnea severity.

As expected, women were over-represented in the baseline AHI < 5 group (64.9%) and men were over-represented in the AHI ³ 15 group (63.5%) (p<0.001). Baseline BMI increased as baseline AHI severity increased.  Overall unadjusted five-year increase in BMI was modest with a mean (SD) BMI change of 0.53 (2.61) kg/m². The unadjusted five-year increase in BMI was 0.63 (2.54) kg/m² for those with baseline AHI < 5, 0.43 (2.48) kg/m² among those with AHI ≥ 5 to < 15 and 0.37 (3.05) kg/m² for the AHI group ≥ 15. These values were not statistically different from each other.

A multivariate regression model was constructed predicting five-year change in BMI by baseline AHI category adjusted for age, gender, race, baseline BMI, and AHI change (Table 2).

Table 2: Adjusted β coefficients of BMI change according to AHI and continuous variables in the Sleep Heart Health Study*.

Compared to baseline AHI group of < 5, those with AHI between ≥ 5 to < 15 had a mean adjusted increase in BMI of 0.21 that approached statistical significance (p=0.055). However, those with AHI ≥ 15 had a statistically significant adjusted BMI increase of 0.51 (p<0.001).  Younger age, lower baseline BMI and greater AHI change also were associated with a larger BMI increase. There was a trend for women to have a greater increase in BMI, but no effect of race was observed. However, the model only accounted for 7% of the total variance. Adjusted means by baseline AHI group are displayed graphically in Figure 1.

Figure 1: Estimated Adjusted Means of BMI increase according to AHI in the Sleep Heart Health Study. Data are adjusted for baseline age (continuous), race (categorical), gender (categorical), baseline BMI (continuous), change in AHI (continuous). Covariates fixed at: baseline BMI = 28.7, baseline age 62.1, change in RDI = 2.7. Bars represent 95% confidence intervals.

Discussion

Our findings indicate that there is a positive association between severity of SDB and five-year increase in BMI. The finding was demonstrated after controlling for key covariates including age, gender, race, baseline BMI, and AHI change. This observation may help explain the difficulty patients with SDB have in trying to lose weight.

Two previous small studies have demonstrated a positive association between newly diagnosed SDB and weight gain. A retrospective study by Phillips et al. compared one-year weight histories of 53 men and women patients who were recently diagnosed with SDB with 24 control subjects matched for gender, age, BMI and percent body fat.⁵ Subjects in that study were somewhat younger than the SHHS cohort with an age difference of approximately 10 years. The SDB among subjects in the previous study tended to be moderate to severe with mean ± SEM AHI 33 ± 5 /h for men and 37 ± 10 /h for women. Mean ± SEM of BMI at time of diagnosis was somewhat higher than in the SHHS with 35 ± 1 kg/m² for men and 44 ± 2 kg/m² for women.  Men and women patients with SDB had reported a recent weight gain of 7.4 ± 1.5 kg compared with a weight loss of 0.5 ± 1.7 kg (p=0.001) in obese controls without SDB. However, given the design of this study it is not possible to determine whether weight gain contributed to the onset of SDB or was a result of SDB. The study was also limited by reliance on self-report of weight gain history.

Another study by Traviss et al. prospectively evaluated 49 obese patients with newly diagnosed SDB.¹⁶ Mean ± SD of AHI at diagnosis was severe at 45 ± 27 /h.  BMI at diagnosis was elevated at 36.5 ± 6.2 kg/m². Of the 49 subjects, 43 could estimate the duration of their symptoms with 84% reporting weight gain since becoming symptomatic.  Weight gain was relatively large, with a reported 17 ± 15 kg over 5.3 ± 4.8 years. However, this study was limited by the lack of a control group and reliance on self-report of weight history.  These two small studies, in addition to our findings, suggest that there is an association between SDB and increased BMI.

Interestingly, unadjusted BMI change in our study was quite modest and not statistically different as a function of SDB severity. However, BMI change over time is a complex phenomenon influenced by several variables. A large (29,799 subjects) prospective study examining 5-year change in weight in a multi-ethnic cohort of men and women explored several of these relationships.¹⁹ In that study, younger men and to a greater degree, younger women were at greater risk for weight gain compared to older adults. This is consistent with our initial findings. In addition, there was a trend for women in the higher baseline BMI categories of ‘overweight’ (BMI >25– 30 kg/m²) and ‘obese’ (BMI >30 kg/m²) in the aforementioned cohort to gain more weight than men in the higher baseline BMI categories. In order to more precisely examine the effect of AHI on weight change, we controlled for these confounders in our final multivariate model thus resulting in the finding of an increase in BMI as a function of SDB severity in this study. 

Several mechanisms could explain why SDB contributes to increased BMI. First, persons with SDB may have a reduction in the quantity and quality of their sleep. Recent data indicate that insufficient sleep may be a risk factor for obesity.²⁰ Experimental sleep restriction increases ghrelin and reduces leptin production favoring appetite enhancement,²¹ a finding that also has been observed in a large population cohort.²² Second, those with SDB may eat a diet that favors weight gain. In support of this hypothesis, sleep restriction has been shown to increase craving for calorie dense food with high carbohydrate content. The Apnea Positive Pressure Long-Term Efficacy Study (APPLES) demonstrated that those with severe SDB consumed a diet higher in cholesterol, protein, total fat and total saturated fatty acids, even after adjusting for BMI, age, and daytime sleepiness.²³ Third, a cardinal symptom of SDB is excessive daytime sleepiness. Thus, it is possible that persons with SDB engage in less physical activity because they are too fatigued to exercise. Data from APPLES indicate that recreational physical activity is less in those with SDB. However, this finding appears to be principally explained by concomitant obesity. 

Weight loss frequently results in an improvement and sometimes resolution in SDB. This is most evident in those who undergo bariatric surgical procedures.^24,25 Persons with SDB are frequently counseled to treat their SDB by losing weight through diet and exercise,²⁶ an approach that is usually unsuccessful.²⁵ Failure to primarily address SDB in conjunction with a weight reduction program may diminish the latter’s success. However, evidence to date indicates that treatment of SDB does not consistently result in weight loss. In a sample of clinical patients with SDB, treatment with CPAP did not result in weight loss. Moreover, in female patients, there was actually an increase in weight.²⁷ In addition, consistent weight reduction was not observed in a small number of patients with severe OSA who underwent tracheostomy.²⁸ Thus, it appears that weight gain engendered by the presence of OSA is not easily reversed despite therapy. Prospective studies will be required to determine whether primary treatment for OSA enhances weight loss programs in those with OSA.

Although this analysis demonstrated a positive association of severity of SDB on five-year increase in BMI, there are several caveats that deserve consideration. The BMI of participants tended to be lower than that seen in clinical SDB populations and a relatively small number of subjects had large changes in BMI. As previously noted, the mean BMI increase was, at best, quite modest. When converted for illustrative purposes to weight using an average height of 167 cm of the participants, those with an AHI between ≥ 5 to < 15 had a mean adjusted increase in BMI of 0.21 kg/m² equal to 0.59 kg or 1.30 lbs. Similarly, those with AHI ≥ 15 had an adjusted BMI increase of 0.51 kg/m² equal to 1.42 kg or 3.13 lbs. Thus, the magnitude of the changes we observed may not be applicable to clinical populations where patients with SDB may have a higher BMI. In addition, it is not known when the participants developed SDB, thus definitive inference of causality cannot be made. However, following a large undiagnosed cohort over an extended period of time to determine incidence of SDB onset and subsequent change in weight would be exceedingly difficult and costly.  Additionally, the model only accounted for a small amount of the total variance in five-year BMI increase, suggesting that there are likely other unmeasured variables influencing the amount of BMI increase over time in this cohort. Finally, while not statistically significant, the unadjusted mean change in BMI was slightly less in the high RDI group in comparison to the lower RDI groups. This observation underscores the biological complexity of the interactions among weight change, SDB, age, gender and other factors.

In conclusion, our findings suggest that although weight gain is a risk factor for developing or worsening SDB, SDB may, in a reciprocal fashion, lead to increased weight gain. This may help explain why patients with SDB find it difficult to lose weight.

Acknowledgements

This work was supported by National Heart, Lung and Blood Institute cooperative agreements U01HL53940 (University of Washington), U01HL53941 (Boston University), U01HL53938 (University of Arizona), U01HL53916 (University of California, Davis), U01HL53934 (University of Minnesota), U01HL53931 (New York University), U01HL53937 and U01HL64360  (Johns Hopkins University), U01HL63463 (Case Western Reserve University), and U01HL63429 (Missouri Breaks Research).

Sleep Heart Health Study (SHHS) acknowledges the Atherosclerosis Risk in Communities Study (ARIC), the Cardiovascular Health Study (CHS), the Framingham Heart Study (FHS), the Cornell/Mt. Sinai Worksite and Hypertension Studies, the Strong Heart Study (SHS), the Tucson Epidemiologic Study of Airways Obstructive Diseases (TES) and the Tucson Health and Environment Study (H&E) for allowing their cohort members to be part of the SHHS and for permitting data acquired by them to be used in the study.  SHHS is particularly grateful to the members of these cohorts who agreed to participate in SHHS as well. SHHS further recognizes all of the investigators and staff who have contributed to its success. A list of SHHS investigators, staff and their participating institutions is available on the SHHS website, www.jhucct.com/shhs.

The opinions expressed in the paper are those of the author(s) and do not necessarily reflect the views of the Indian Health Service.

These data have been presented in part at the Annual Meeting of the Associated Professional Sleep Societies, June 11, 2009, Seattle, WA.

References

1. Bixler EO, Vgontzas AN, Lin HM, Ten Have T, Rein J, Vela-Bueno A, et al. Prevalence of sleep-disordered breathing in women: Effects of gender. Am J Respir Crit Care Med 2001;163:608-13.
2. Bixler EO, Vgontzas AN, Ten Have T, Tyson K, Kales A. Effects of age on sleep apnea in men: I. prevalence and severity. Am J Respir Crit Care Med 1998;157:144-8.
3. Carmelli D, Swan GE, Bliwise DL. Relationship of 30-year changes in obesity to sleep-disordered breathing in the western collaborative group study. Obes Res 2000;8:632-7.
4. Duran J, Esnaola S, Rubio R, Iztueta A. Obstructive sleep apnea-hypopnea and related clinical features in a population-based sample of subjects aged 30 to 70 yr. Am J Respir Crit Care Med 2001;163:685-9.
5. Phillips BG, Hisel TM, Kato M, Pesek CA, Dyken ME, Narkiewicz K, et al. Recent weight gain in patients with newly diagnosed obstructive sleep apnea. J Hypertens 1999;17:1297-300.
6. Tishler PV, Larkin EK, Schluchter MD, Redline S. Incidence of sleep-disordered breathing in an urban adult population: The relative importance of risk factors in the development of sleep-disordered breathing. JAMA 2003;289:2230-7.
7. Young T, Shahar E, Nieto FJ, Redline S, Newman AB, Gottlieb DJ, et al. Predictors of sleep-disordered breathing in community-dwelling adults: The sleep heart health study. Arch Intern Med 2002;162:893-900.
8. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328:1230-5.
9. Young T, Peppard PE, Taheri S. Excess weight and sleep-disordered breathing. J Appl Physiol 2005;99:1592-9.
10. Newman AB, Foster G, Givelber R, Nieto FJ, Redline S, Young T. Progression and regression of sleep-disordered breathing with changes in weight: The sleep heart health study. Arch Intern Med 2005;165:2408-13.
11. Ip MS, Lam B, Tang LC, Lauder IJ, Ip TY, Lam WK. A community study of sleep-disordered breathing in middle-aged chinese women in hong kong: Prevalence and gender differences. Chest 2004;125:127-34.
12. Millman RP, Carlisle CC, McGarvey ST, Eveloff SE, Levinson PD. Body fat distribution and sleep apnea severity in women. Chest 1995;107:362-6.
13. Barvaux VA, Aubert G, Rodenstein DO. Weight loss as a treatment for obstructive sleep apnoea. Sleep Med Rev 2000;4:435-52.
14. Grunstein RR, Stenlof K, Hedner JA, Peltonen M, Karason K, Sjostrom L. Two year reduction in sleep apnea symptoms and associated diabetes incidence after weight loss in severe obesity. Sleep 2007;30:703-10.
15. Peppard PE, Young T, Palta M, Dempsey J, Skatrud J. Longitudinal study of moderate weight change and sleep-disordered breathing. JAMA 2000;284:3015-21.
16. Traviss KA, Barr SI, Fleming JA, Ryan CF. Lifestyle-related weight gain in obese men with newly diagnosed obstructive sleep apnea. J Am Diet Assoc 2002;102:703-6.
17. Quan SF, Howard BV, Iber C, Kiley JP, Nieto FJ, O'Connor GT, et al. The sleep heart health study: Design, rationale, and methods. Sleep 1997;20:1077-85.
18. Redline S, Sanders MH, Lind BK, Quan SF, Iber C, Gottlieb DJ, et al. Methods for obtaining and analyzing unattended polysomnography data for a multicenter study. sleep heart health research group. Sleep1998;21:759-67.
19. Ball K, Crawford D, Ireland P, Hodge A. Patterns and demographic predictors of 5-year weight change in a multi-ethnic cohort of men and women in australia. Public Health Nutr 2003;6:269-81.
20. Chaput JP, Despres JP, Bouchard C, Tremblay A. The association between sleep duration and weight gain in adults: A 6-year prospective study from the Quebec Family Study. Sleep 2008;31:517-23.
21. Spiegel K, Tasali E, Penev P, Van Cauter E. Brief communication: Sleep curtailment in healthy young men is associated with decreased leptin levels, elevated ghrelin levels, and increased hunger and appetite. Ann Intern Med 2004;141:846-50.
22. Taheri S, Lin L, Austin D, Young T, Mignot E. Short sleep duration is associated with reduced leptin, elevated ghrelin, and increased body mass index. PLoS Med. 2004;1:e62.
23. Vasquez MM, Goodwin JL, Drescher AA, Smith TW, Quan SF. Associations of dietary intake and physical activity with sleep disordered breathing in the apnea positive pressure long-term efficacy study (APPLES). J Clin Sleep Med 2008;4:411-8.
24. Rasheid S, Banasiak M, Gallagher SF, Lipska A, Kaba S, Ventimiglia D, et al. Gastric bypass is an effective treatment for obstructive sleep apnea in patients with clinically significant obesity. Obes Surg 2003;13:58-61.
25. Fritscher LG, Canani S, Mottin CC, Fritscher CC, Berleze D, Chapman K, et al. Bariatric surgery in the treatment of obstructive sleep apnea in morbidly obese patients. Respiration 2007;74:647-52.
26. Veasey SC, Guilleminault C, Strohl KP, Sanders MH, Ballard RD, Magalang UJ. Medical therapy for obstructive sleep apnea: A review by the medical therapy for obstructive sleep apnea task force of the standards of practice committee of the american academy of sleep medicine. Sleep 2006;29:1036-44.
27. Redenius R, Murphy C, O'Neill E, Al-Hamwi M, Zallek SN. Does CPAP lead to change in BMI? J Clin Sleep Med 2008;4:205-9.
28. Haapaniemi JJ, Laurikainen EA, Halme P, Antila J. Long-term results of tracheostomy for severe obstructive sleep apnea syndrome. ORL J Otorhinolaryngol Relat Spec 2001;63:131-6.