CLINICAL INVESTIGATION
Effect of Obstructive Sleep Apnea Treatment on Mail-In Cognitive Function Screening Instrument Brett C. Bade, MD, Charlie Strange, MD and Chitra Lal, MD
Abstract: Background: Obstructive sleep apnea syndrome (OSAS) may be associated with cognitive impairment (CI). The goal of this study was to evaluate the impact of risk factors and continuous positive airway pressure (CPAP) on a screening tool for cognitive function. Methods: The Mail-In Cognitive Function Screening Instrument (MCFSI) is a self-administered test designed to identify CI in the Alzheimer’s Disease Cooperative Study. It was administered to 88 consecutive patients with OSAS attending the Medical University of South Carolina Sleep Clinic. An MCFSI score $5 was considered abnormal. Results: Data were analyzed on 61 patients after excluding missing and duplicate data. The MCFSI score was abnormal in 15 patients (25%). African Americans were more likely to be CPAP-noncompliant. Female gender and smoking were associated with abnormal MCFSI scores. CPAP-compliant patients were more likely to have normal MCFSI scores, although the difference was not statistically significant (P 5 0.06). Conclusions: CPAP-compliant patients showed a trend toward lower MCFSI scores. There may be gender and racial differences in CI related to OSAS, predisposing certain groups to worse morbidity. Appropriate treatment and compliance with CPAP could improve CI in OSAS. Larger studies with multivariate analyses are needed to identify relationships between individual OSAS and CI risk factors. Key Indexing Terms: Sleep apnea; Cognitive impairment; Continuous positive airway pressure; Risk factors; Mail-in cognitive function screening instrument. [Am J Med Sci 2014;348(3):215–218.]
O
bstructive sleep apnea syndrome (OSAS) is a common disorder with a high prevalence in the general population.1 It is characterized by a decrease or cessation of normal airflow during sleep, which is often accompanied by symptoms of excessive daytime sleepiness (EDS), fatigue, cardiovascular2 and metabolic sequelae.3 As the prevalence of OSAS rises in an aging population with increasing obesity, data about its adverse impact on health continue to emerge. There is increasing evidence that OSAS is associated with impaired cognitive functioning.4 The cognitive domains that are most commonly affected by OSAS include vigilance,5 executive function6 and memory.7 One important aspect of OSAS-associated cognitive impairment (CI) is that some populations may be particularly at risk. Therefore, screening tools From the Departments of Internal Medicine (BCB) and Pulmonary, Critical Care, Allergy and Sleep Medicine (CS, CL), Medical University of South Carolina (MUSC), Charleston, South Carolina. Submitted July 23, 2013; accepted in revised form November 18, 2013. C. Lal has received grants from South Carolina Clinical and translational research program and Invado Pharmaceuticals for sleep apnea research and has also received travel reimbursement from University of Colorado and St. Joseph Hospital, Chicago, Illinois, for CME presentations as a guest speaker. The remaining authors have no financial or other conflicts of interest to disclose. Presented at the Annual American Thoracic Society Meeting, May 20, 2013, Philadelphia, PA. Correspondence: Chitra Lal, MD, Department of Pulmonary, Critical Care, Allergy and Sleep Medicine, Medical University of South Carolina (MUSC), 96 Jonathan Lucas Street, CSB 812, Msc 630, Charleston, SC 29425 (E-mail: [email protected]).
The American Journal of the Medical Sciences
that target populations such as the elderly8 may allow improved clinical recognition of the disease state. Given the general irreversibility of most dementias, it is critical that any reversible etiologies for CI should be identified and treated. Objective neurocognitive assessment requires intensive testing and trained personnel, both of which are time consuming and expensive. The Mail-In Cognitive Function Screening Instrument (MCFSI) was developed for the Alzheimer’s Disease Cooperative Study Prevention Instrument Project as an open-access survey to evaluate whether a brief screening tool could be used to trigger a diagnostic evaluation in large dementia prevention trials.9 The MCFSI is a short, self-administered, 14-point test meant to detect subjective memory impairment in nondemented individuals. Importantly, the MCFSI measures the degree of self-perceived CI with higher scores correlating with worse cognition. A correlation has been seen between the MCFSI total scores, the Mini-Mental Status Examination scores and the Clinical Dementia Rating Scores in healthy elderly individuals. We have adopted the MCFSI as a quick and easy to administer screening tool for CI in our OSAS population. To assess the impact of continuous positive airway pressure (CPAP) treatment of OSAS on cognitive function, we evaluated the MCFSI scores in CPAP-compliant and noncompliant patients. We postulated that MCFSI scores would be lower in CPAP-compliant patients, thus suggesting evidence of better cognitive function with CPAP treatment.
METHODS After approval from the Medical University of South Carolina’s Institutional Review Board, waiver of informed consent was allowed to collect retrospective data on 88 sequential patients presenting to the Sleep Clinic from January 1, 2012 to August 30, 2012. Before seeing their physician for a clinic visit, patients were administered the MCFSI questionnaire. The components of the MCFSI are shown in Table 1. In addition to demographics, we collected data regarding known risk factors for OSAS and neurocognitive dysfunction. These risk factors included obesity, smoking and alcohol abuse, hypertension, diabetes mellitus, metabolic syndrome and use of psychoactive medications. Severity of OSAS was evaluated using the apnea hypopnea index (AHI) or respiratory disturbance index (RDI) from the most recent overnight polysomnogram. Patients with (1) an AHI or RDI $5/hr with symptoms or (2) an AHI or RDI $15/hr with/without symptoms were considered to have OSAS. Data on the oxygen desaturation index were not collected. CPAP compliance was defined as usage of CPAP for $4 hours per night for 70% of nights in a consecutive 30-day period within the past 3 months, in accordance with the Medicare definition of CPAP compliance. Comparative data were analyzed using the 2-tailed Student t test for normally distributed continuous data and x2 analysis with Pearson correlation coefficient for categorical data (JMP, Cary, NC). MCSFI was not normally distributed; therefore, a Wilcoxon rank sums test was used for analysis. CPAP compliance was compared with demographics, AHI/RDI and
Volume 348, Number 3, September 2014
215
Bade et al
TABLE 1. Mail-in cognitive function screening instrument components Yes 5 1
Question
No 5 0
Maybe 5 0.5
1. Compared to 1 year ago, do you feel that your memory has declined substantially? 2. Do others tell you that you tend to repeat questions over and over? 3. Have you been misplacing things more often? 4. Do you find that lately you are relying more on written reminders? 5. Do you need more help from others to remember appointments, family occasions or holidays? 6. Do you have more trouble recalling names, finding the right word or completing sentences? 7. Do you have more trouble driving? 8. Compared to 1 year ago, do you have more difficulty managing money? 9. Are you less involved in social activities? 10. Has your work performance declined significantly compared to 1 year ago? 11. Do you have more trouble following the news, or the plots of books, movies or TV shows, compared to 1 year ago? 12. Are there any activities that are substantially more difficult for you now compared to 1 year ago? 13. Are you more likely to become disorientated, or get lost, for example when traveling to another city? 14. Do you have more difficulty using household appliances? Total score
MCFSI scores. MCFSI scores (normal versus abnormal) were compared with demographics, AHI/RDI (whichever was greater) and known CI risk factors. P values of #0.05 were considered significant. MCFSI scores $5 were considered abnormal for this study. The MCFSI has not been studied enough to have firmly established normal and abnormal values. However, the mean MCFSI in Alzheimer’s Disease Cooperative Study subjects was 2.4. We chose values $5 to try to ensure significant results.
RESULTS The Consort diagram is shown in Figure 1. A complete data set was available on 61 patients with OSAS. Twenty-seven patients were excluded because of duplicate visits or incomplete data acquisition. The individuals’ demographics and test data are presented in Table 2. Average duration of CPAP use in the CPAP-compliant group was $6 weeks.
Age was similar in the CPAP-compliant and noncompliant groups and in subjects with normal and abnormal MCFSI scores. African American ethnicity was associated with a greater likelihood of CPAP noncompliance. In addition, African Americans and women were more likely to have abnormal MCFSI scores. These results were not adjusted for education level. The CPAP-compliant group had worse OSAS at baseline as defined by the AHI/RDI. The MCFSI was abnormal ($5) in 15 patients (25%). A greater percentage of CPAP-compliant patients had normal MCFSI scores as compared with the CPAP-noncompliant group, with a P value nearing significance (P 5 0.06). The correlation between individual MCFSI scores and CPAP compliance is shown in Figure 2. To further understand the relationship between OSAS and MCFSI scores, multiple univariate analyses were performed (Table 3). In addition to the gender and race differences described above, smoking was associated with abnormal MCFSI scores (P 5 0.01). Use of psychoactive medications showed a trend towards abnormal MCFSI scores (P 5 0.06).
DISCUSSION
FIGURE 1.
216
Consort diagram of subject inclusion and data analysis.
This study shows a strong correlation between CPAP compliance and better cognitive functioning as measured by the MCFSI questionnaire. This proof of principal study was needed to assist in powering future studies and assuring the instrument was easy to use. Our real world effectiveness evaluation was helpful in the clinic and correlated in an anecdotal manner with reports of family and friends who accompanied patients to the clinic. As we had expected, MCFSI scores are not normally distributed in the OSAS population. Most individuals score in the normal range, making screening research particularly difficult in this disease. There are several potential explanations for MCFSI variability. Because MCFSI is a self-perception of cognition, we would not expect all patients to recognize their CI due to insidious onset, mild impairment or potential denial. Variability also might occur day-to-day because impaired vigilance from EDS also impacts cognitive functioning. Volume 348, Number 3, September 2014
Sleep Apnea and Memory
TABLE 2. Baseline demographic characteristics of CPAPcompliant versus noncompliant patients CPAPCPAPcompliant noncompliant (n 5 37) (n 5 24) P Age Race (% African American) Sex (% female) Baseline RDI or AHIa MCFSI mean % Normal (,5) Mean MCFSI score if abnormal
55.1 6 11.7 40.5
54.6 6 12.1 75.0
0.44 0.01
67.6 35.5 6 28.7 2.2 6 2.7 83.78 7.0 6 2.3
75.0 23.6 6 16.2 3.9 6 3.8 62.5 8.1 6 2.3
0.53 0.02 0.10 0.06 0.80
a Baseline RDI or AHI was measured on the initial diagnostic sleep study before therapy. AHI, apnea hypopnea index; RDI, respiratory disturbance index; MCFSI, Mail-In Cognitive Function Screening Instrument.
There was a trend towards lower MCFSI scores in CPAP-compliant patients and higher scores in CPAPnoncompliant patients. Because of the frequency of normal MCFSI reports (,5), we evaluated the percentage of normal responses in the CPAP-compliant versus CPAP-noncompliant groups. We found that 84% of CPAP-compliant patients had a normal MCFSI score, whereas only 63% of noncompliant patients had a normal MCFSI (P 5 0.06). Although we have not accounted for treatment related bias (ie, CPAP-compliant patients may feel that they are better as compared with CPAP-noncompliant patients), it is important to note that MCFSI was designed to identify patients with selfperceived impairment. Self-identification seems better than observer’s identification of impairment.9 Thus, first, we know that some patients can identify their own impairment, and our study relies on their perception of impairment and improvement.
FIGURE 2. Correlation between MCFSI scores and CPAP compliance. Individual dots represent MCFSI scores of each individual subject. The blue line in the middle represents the median cutoff value for MCFSI scores in each group. The P value of 0.06 is for the likelihood of having a normal MCFSI score (,5) in CPAP-compliant versus CPAP-noncompliant patients. Ó 2014 Lippincott Williams & Wilkins
TABLE 3. Correlation between MCFSI scores and risk factors for cognitive impairment Normal Abnormal MCFSI MCFSI (‡5) (n 5 46) (n 5 15) P Age Race (% African American) Sex (% female) RDI or AHI Obesity (BMI $30), % Metabolic syndrome, % Smoking, % Psychoactive medications, % CHF, % Alcohol, %
55.5 6 11.6 43.5
52.9 6 12.2 86.7
0.44 ,0.01
60.9 34.1 6 26.7 67.5 52.2 2.2 45.7
100 20.8 6 16.2 42.9 33.3 20.0 73.3
,0.01 0.08 0.25 0.20 0.01 0.06
8.7 45.7
26.7 26.7
0.07 0.19
AHI, apnea hypopnea index; RDI, respiratory disturbance index; MCFSI, Mail-In Cognitive Function Screening Instrument; BMI, body mass index; CHF, congestive heart failure.
Second, we know that CPAP improves EDS and alertness level; improvement in alertness may make patients with OSAS more likely to recognize their impairment. In that case, treatment would not represent a bias, but in fact proof of improvement. Third, in our cohort, CPAP-compliant patients had a higher baseline RDI/AHI, suggesting patients with worse disease recognized some type of improvement (likely in EDS) and thus were more likely to be compliant with CPAP. The duration of OSAS could impact the severity and/or reversibility of CI. This concern is heightened by the often occult nature of OSAS and the frequent delay in diagnosis. A recent study has shown improvements in memory, attention, executive function and gray-matter volume on neuroimaging after 3 months of CPAP therapy in a small cohort of patients with OSAS.10 Our preliminary data further suggest that appropriate recognition, diagnosis, treatment and compliance with CPAP might improve subjective memory and cognitive complaints in patients with OSAS. We suspect that the MCFSI may be a more sensitive tool for early recognition of mild cognitive impairment than patient and family self-reports. Memory impairment is relatively advanced by the time it is recognized by patients or their family members as anecdotally noted in more advanced dementia patients. If true, this would support the use of a sensitive screening tool to facilitate early diagnosis and aggressive treatment of OSAS with CPAP. Consistent with previous studies, smoking and use of psychoactive medications (eg, antidepressants, mood stabilizers and anxiolytics) were associated with worse cognition as measured by the MCFSI. Unexpectedly, age, severity of AHI/ RDI, body mass index, presence of metabolic syndrome, presence of congestive heart failure and alcohol use did not correlate with elevated MCFSI scores. We had presumed that risk factors associated with CI would worsen MCFSI scores in an additive/synergistic fashion. These types of questions are better answered by larger studies that generate higher power and allow multivariate analysis. African Americans and women were found to have higher MCFSI scores (P , 0.01 for both). Although these associations could be a type I error, further study of the MCFSI
217
Bade et al
in OSAS will require race- and gender-specific validation. Therefore, the MCFSI may require modification to avoid any intrinsic bias related to question formulation or cultural sensitivity. Also, differences in education level, which were not collected in the study cohort, could influence MCFSI scores. It is certainly possible that race, gender and educational level could have an actual impact on cognitive function. There was no statistically significant difference in the AHI/RDI between patients with normal versus abnormal MCFSI scores. The MCFSI scores were collected on patients with OSAS who were already on CPAP treatment, hence any correlation between MCFSI scores and AHI/RDI could not be established in this study. However, it is possible that patients with a higher AHI/RDI were more likely to use CPAP (as seen by the higher AHI/RDI in the CPAP-compliant group) and thus more likely to have normal MCFSI scores, as seen by the trend for higher AHI/RDI in the group with normal MCFSI scores. Conversely, it is also possible that the higher level of CI in CPAP-noncompliant patients may itself have contributed to less use of CPAP, thus creating a vicious cycle of worsening CI and other sequelae of OSAS. There are several limitations to this study. These include a small cohort size and lack of formal neuropsychological testing. We did not have enough power to establish the strength of individual risk factors in the causation of CI in OSAS subjects. Data on the oxygen desaturation index were not collected. Also, the relationship between OSAS and cognitive function may not be linear (possibly because of complex relationships between risk factors and disease). Future large studies are needed to further evaluate the association between OSAS and CI, the pathophysiological mechanisms underlying the CI, potential race/gender associations and targeted strategies to prevent/reverse the process. Further studies to evaluate the relationships between severity, gender and race are necessary. To the best of our knowledge, our study is the first one to use the MCFSI questionnaire for evaluation of cognitive function in patients with OSAS. Larger studies are needed to further validate the utility of this quick and simple screening tool for assessment of cognitive function in the OSAS population.
218
CONCLUSIONS CPAP compliance may improve the CI associated with OSAS. Thus, prompt diagnosis and treatment of OSAS and encouragement of CPAP use in patients with OSAS are critical. CI in OSAS is influenced by several factors and these relationships need to be evaluated further in future large studies. REFERENCES 1. Young T, Palta M, Dempsey J, et al. The occurrence of sleepdisordered breathing among middle-aged adults. N Engl J Med 1993; 328:1230–5. 2. Fava C, Montagnana M, Favaloro EJ, et al. Obstructive sleep apnea syndrome and cardiovascular diseases. Semin Thromb Hemost 2011;37: 280–97. 3. Fusetti M, Fioretti AB, Valenti M, et al. Cardiovascular and metabolic comorbidities in patients with obstructive sleep apnoea syndrome. Acta Otorhinolaryngol Ital 2012;32:320–5. 4. Lal C, Strange C, Bachman D. Neurocognitive impairment in obstructive sleep apnea. Chest 2012;141:1601–10. 5. Findley L, Unverzagt M, Guchu R, et al. Vigilance and automobile accidents in patients with sleep apnea or narcolepsy. Chest 1995;108: 619–24. 6. Naismith S, Winter V, Gotsopoulos H, et al. Neurobehavioral functioning in obstructive sleep apnea: differential effects of sleep quality, hypoxemia and subjective sleepiness. J Clin Exp Neuropsychol 2004; 26:43–54. 7. Wallace A, Bucks RS. Memory and obstructive sleep apnea: a metaanalysis. Sleep 2013;36:203–20. 8. Kim SJ, Lee JH, Lee DY, et al. Neurocognitive dysfunction associated with sleep quality and sleep apnea in patients with mild cognitive impairment. Am J Geriatr Psychiatry 2011;19:374–81. 9. Walsh SP, Raman R, Jones KB, et al. ADCS Prevention Instrument Project: the Mail-In Cognitive Function Screening Instrument (MCFSI). Alzheimer Dis Assoc Disord 2006;20:S170–8. 10. Canessa N, Castronovo V, Cappa SF, et al. Obstructive sleep apnea: brain structural changes and neurocognitive function before and after treatment. Am J Respir Crit Care Med 2011;183:1419–26.
Volume 348, Number 3, September 2014
Effect of Obstructive Sleep Apnea Treatment on Mail-In Cognitive Function Screening Instrument Brett C. Bade, MD, Charlie Strange, MD and Chitra Lal, MD
Abstract: Background: Obstructive sleep apnea syndrome (OSAS) may be associated with cognitive impairment (CI). The goal of this study was to evaluate the impact of risk factors and continuous positive airway pressure (CPAP) on a screening tool for cognitive function. Methods: The Mail-In Cognitive Function Screening Instrument (MCFSI) is a self-administered test designed to identify CI in the Alzheimer’s Disease Cooperative Study. It was administered to 88 consecutive patients with OSAS attending the Medical University of South Carolina Sleep Clinic. An MCFSI score $5 was considered abnormal. Results: Data were analyzed on 61 patients after excluding missing and duplicate data. The MCFSI score was abnormal in 15 patients (25%). African Americans were more likely to be CPAP-noncompliant. Female gender and smoking were associated with abnormal MCFSI scores. CPAP-compliant patients were more likely to have normal MCFSI scores, although the difference was not statistically significant (P 5 0.06). Conclusions: CPAP-compliant patients showed a trend toward lower MCFSI scores. There may be gender and racial differences in CI related to OSAS, predisposing certain groups to worse morbidity. Appropriate treatment and compliance with CPAP could improve CI in OSAS. Larger studies with multivariate analyses are needed to identify relationships between individual OSAS and CI risk factors. Key Indexing Terms: Sleep apnea; Cognitive impairment; Continuous positive airway pressure; Risk factors; Mail-in cognitive function screening instrument. [Am J Med Sci 2014;348(3):215–218.]
O
bstructive sleep apnea syndrome (OSAS) is a common disorder with a high prevalence in the general population.1 It is characterized by a decrease or cessation of normal airflow during sleep, which is often accompanied by symptoms of excessive daytime sleepiness (EDS), fatigue, cardiovascular2 and metabolic sequelae.3 As the prevalence of OSAS rises in an aging population with increasing obesity, data about its adverse impact on health continue to emerge. There is increasing evidence that OSAS is associated with impaired cognitive functioning.4 The cognitive domains that are most commonly affected by OSAS include vigilance,5 executive function6 and memory.7 One important aspect of OSAS-associated cognitive impairment (CI) is that some populations may be particularly at risk. Therefore, screening tools From the Departments of Internal Medicine (BCB) and Pulmonary, Critical Care, Allergy and Sleep Medicine (CS, CL), Medical University of South Carolina (MUSC), Charleston, South Carolina. Submitted July 23, 2013; accepted in revised form November 18, 2013. C. Lal has received grants from South Carolina Clinical and translational research program and Invado Pharmaceuticals for sleep apnea research and has also received travel reimbursement from University of Colorado and St. Joseph Hospital, Chicago, Illinois, for CME presentations as a guest speaker. The remaining authors have no financial or other conflicts of interest to disclose. Presented at the Annual American Thoracic Society Meeting, May 20, 2013, Philadelphia, PA. Correspondence: Chitra Lal, MD, Department of Pulmonary, Critical Care, Allergy and Sleep Medicine, Medical University of South Carolina (MUSC), 96 Jonathan Lucas Street, CSB 812, Msc 630, Charleston, SC 29425 (E-mail: [email protected]).
The American Journal of the Medical Sciences
that target populations such as the elderly8 may allow improved clinical recognition of the disease state. Given the general irreversibility of most dementias, it is critical that any reversible etiologies for CI should be identified and treated. Objective neurocognitive assessment requires intensive testing and trained personnel, both of which are time consuming and expensive. The Mail-In Cognitive Function Screening Instrument (MCFSI) was developed for the Alzheimer’s Disease Cooperative Study Prevention Instrument Project as an open-access survey to evaluate whether a brief screening tool could be used to trigger a diagnostic evaluation in large dementia prevention trials.9 The MCFSI is a short, self-administered, 14-point test meant to detect subjective memory impairment in nondemented individuals. Importantly, the MCFSI measures the degree of self-perceived CI with higher scores correlating with worse cognition. A correlation has been seen between the MCFSI total scores, the Mini-Mental Status Examination scores and the Clinical Dementia Rating Scores in healthy elderly individuals. We have adopted the MCFSI as a quick and easy to administer screening tool for CI in our OSAS population. To assess the impact of continuous positive airway pressure (CPAP) treatment of OSAS on cognitive function, we evaluated the MCFSI scores in CPAP-compliant and noncompliant patients. We postulated that MCFSI scores would be lower in CPAP-compliant patients, thus suggesting evidence of better cognitive function with CPAP treatment.
METHODS After approval from the Medical University of South Carolina’s Institutional Review Board, waiver of informed consent was allowed to collect retrospective data on 88 sequential patients presenting to the Sleep Clinic from January 1, 2012 to August 30, 2012. Before seeing their physician for a clinic visit, patients were administered the MCFSI questionnaire. The components of the MCFSI are shown in Table 1. In addition to demographics, we collected data regarding known risk factors for OSAS and neurocognitive dysfunction. These risk factors included obesity, smoking and alcohol abuse, hypertension, diabetes mellitus, metabolic syndrome and use of psychoactive medications. Severity of OSAS was evaluated using the apnea hypopnea index (AHI) or respiratory disturbance index (RDI) from the most recent overnight polysomnogram. Patients with (1) an AHI or RDI $5/hr with symptoms or (2) an AHI or RDI $15/hr with/without symptoms were considered to have OSAS. Data on the oxygen desaturation index were not collected. CPAP compliance was defined as usage of CPAP for $4 hours per night for 70% of nights in a consecutive 30-day period within the past 3 months, in accordance with the Medicare definition of CPAP compliance. Comparative data were analyzed using the 2-tailed Student t test for normally distributed continuous data and x2 analysis with Pearson correlation coefficient for categorical data (JMP, Cary, NC). MCSFI was not normally distributed; therefore, a Wilcoxon rank sums test was used for analysis. CPAP compliance was compared with demographics, AHI/RDI and
Volume 348, Number 3, September 2014
215
Bade et al
TABLE 1. Mail-in cognitive function screening instrument components Yes 5 1
Question
No 5 0
Maybe 5 0.5
1. Compared to 1 year ago, do you feel that your memory has declined substantially? 2. Do others tell you that you tend to repeat questions over and over? 3. Have you been misplacing things more often? 4. Do you find that lately you are relying more on written reminders? 5. Do you need more help from others to remember appointments, family occasions or holidays? 6. Do you have more trouble recalling names, finding the right word or completing sentences? 7. Do you have more trouble driving? 8. Compared to 1 year ago, do you have more difficulty managing money? 9. Are you less involved in social activities? 10. Has your work performance declined significantly compared to 1 year ago? 11. Do you have more trouble following the news, or the plots of books, movies or TV shows, compared to 1 year ago? 12. Are there any activities that are substantially more difficult for you now compared to 1 year ago? 13. Are you more likely to become disorientated, or get lost, for example when traveling to another city? 14. Do you have more difficulty using household appliances? Total score
MCFSI scores. MCFSI scores (normal versus abnormal) were compared with demographics, AHI/RDI (whichever was greater) and known CI risk factors. P values of #0.05 were considered significant. MCFSI scores $5 were considered abnormal for this study. The MCFSI has not been studied enough to have firmly established normal and abnormal values. However, the mean MCFSI in Alzheimer’s Disease Cooperative Study subjects was 2.4. We chose values $5 to try to ensure significant results.
RESULTS The Consort diagram is shown in Figure 1. A complete data set was available on 61 patients with OSAS. Twenty-seven patients were excluded because of duplicate visits or incomplete data acquisition. The individuals’ demographics and test data are presented in Table 2. Average duration of CPAP use in the CPAP-compliant group was $6 weeks.
Age was similar in the CPAP-compliant and noncompliant groups and in subjects with normal and abnormal MCFSI scores. African American ethnicity was associated with a greater likelihood of CPAP noncompliance. In addition, African Americans and women were more likely to have abnormal MCFSI scores. These results were not adjusted for education level. The CPAP-compliant group had worse OSAS at baseline as defined by the AHI/RDI. The MCFSI was abnormal ($5) in 15 patients (25%). A greater percentage of CPAP-compliant patients had normal MCFSI scores as compared with the CPAP-noncompliant group, with a P value nearing significance (P 5 0.06). The correlation between individual MCFSI scores and CPAP compliance is shown in Figure 2. To further understand the relationship between OSAS and MCFSI scores, multiple univariate analyses were performed (Table 3). In addition to the gender and race differences described above, smoking was associated with abnormal MCFSI scores (P 5 0.01). Use of psychoactive medications showed a trend towards abnormal MCFSI scores (P 5 0.06).
DISCUSSION
FIGURE 1.
216
Consort diagram of subject inclusion and data analysis.
This study shows a strong correlation between CPAP compliance and better cognitive functioning as measured by the MCFSI questionnaire. This proof of principal study was needed to assist in powering future studies and assuring the instrument was easy to use. Our real world effectiveness evaluation was helpful in the clinic and correlated in an anecdotal manner with reports of family and friends who accompanied patients to the clinic. As we had expected, MCFSI scores are not normally distributed in the OSAS population. Most individuals score in the normal range, making screening research particularly difficult in this disease. There are several potential explanations for MCFSI variability. Because MCFSI is a self-perception of cognition, we would not expect all patients to recognize their CI due to insidious onset, mild impairment or potential denial. Variability also might occur day-to-day because impaired vigilance from EDS also impacts cognitive functioning. Volume 348, Number 3, September 2014
Sleep Apnea and Memory
TABLE 2. Baseline demographic characteristics of CPAPcompliant versus noncompliant patients CPAPCPAPcompliant noncompliant (n 5 37) (n 5 24) P Age Race (% African American) Sex (% female) Baseline RDI or AHIa MCFSI mean % Normal (,5) Mean MCFSI score if abnormal
55.1 6 11.7 40.5
54.6 6 12.1 75.0
0.44 0.01
67.6 35.5 6 28.7 2.2 6 2.7 83.78 7.0 6 2.3
75.0 23.6 6 16.2 3.9 6 3.8 62.5 8.1 6 2.3
0.53 0.02 0.10 0.06 0.80
a Baseline RDI or AHI was measured on the initial diagnostic sleep study before therapy. AHI, apnea hypopnea index; RDI, respiratory disturbance index; MCFSI, Mail-In Cognitive Function Screening Instrument.
There was a trend towards lower MCFSI scores in CPAP-compliant patients and higher scores in CPAPnoncompliant patients. Because of the frequency of normal MCFSI reports (,5), we evaluated the percentage of normal responses in the CPAP-compliant versus CPAP-noncompliant groups. We found that 84% of CPAP-compliant patients had a normal MCFSI score, whereas only 63% of noncompliant patients had a normal MCFSI (P 5 0.06). Although we have not accounted for treatment related bias (ie, CPAP-compliant patients may feel that they are better as compared with CPAP-noncompliant patients), it is important to note that MCFSI was designed to identify patients with selfperceived impairment. Self-identification seems better than observer’s identification of impairment.9 Thus, first, we know that some patients can identify their own impairment, and our study relies on their perception of impairment and improvement.
FIGURE 2. Correlation between MCFSI scores and CPAP compliance. Individual dots represent MCFSI scores of each individual subject. The blue line in the middle represents the median cutoff value for MCFSI scores in each group. The P value of 0.06 is for the likelihood of having a normal MCFSI score (,5) in CPAP-compliant versus CPAP-noncompliant patients. Ó 2014 Lippincott Williams & Wilkins
TABLE 3. Correlation between MCFSI scores and risk factors for cognitive impairment Normal Abnormal MCFSI MCFSI (‡5) (n 5 46) (n 5 15) P Age Race (% African American) Sex (% female) RDI or AHI Obesity (BMI $30), % Metabolic syndrome, % Smoking, % Psychoactive medications, % CHF, % Alcohol, %
55.5 6 11.6 43.5
52.9 6 12.2 86.7
0.44 ,0.01
60.9 34.1 6 26.7 67.5 52.2 2.2 45.7
100 20.8 6 16.2 42.9 33.3 20.0 73.3
,0.01 0.08 0.25 0.20 0.01 0.06
8.7 45.7
26.7 26.7
0.07 0.19
AHI, apnea hypopnea index; RDI, respiratory disturbance index; MCFSI, Mail-In Cognitive Function Screening Instrument; BMI, body mass index; CHF, congestive heart failure.
Second, we know that CPAP improves EDS and alertness level; improvement in alertness may make patients with OSAS more likely to recognize their impairment. In that case, treatment would not represent a bias, but in fact proof of improvement. Third, in our cohort, CPAP-compliant patients had a higher baseline RDI/AHI, suggesting patients with worse disease recognized some type of improvement (likely in EDS) and thus were more likely to be compliant with CPAP. The duration of OSAS could impact the severity and/or reversibility of CI. This concern is heightened by the often occult nature of OSAS and the frequent delay in diagnosis. A recent study has shown improvements in memory, attention, executive function and gray-matter volume on neuroimaging after 3 months of CPAP therapy in a small cohort of patients with OSAS.10 Our preliminary data further suggest that appropriate recognition, diagnosis, treatment and compliance with CPAP might improve subjective memory and cognitive complaints in patients with OSAS. We suspect that the MCFSI may be a more sensitive tool for early recognition of mild cognitive impairment than patient and family self-reports. Memory impairment is relatively advanced by the time it is recognized by patients or their family members as anecdotally noted in more advanced dementia patients. If true, this would support the use of a sensitive screening tool to facilitate early diagnosis and aggressive treatment of OSAS with CPAP. Consistent with previous studies, smoking and use of psychoactive medications (eg, antidepressants, mood stabilizers and anxiolytics) were associated with worse cognition as measured by the MCFSI. Unexpectedly, age, severity of AHI/ RDI, body mass index, presence of metabolic syndrome, presence of congestive heart failure and alcohol use did not correlate with elevated MCFSI scores. We had presumed that risk factors associated with CI would worsen MCFSI scores in an additive/synergistic fashion. These types of questions are better answered by larger studies that generate higher power and allow multivariate analysis. African Americans and women were found to have higher MCFSI scores (P , 0.01 for both). Although these associations could be a type I error, further study of the MCFSI
217
Bade et al
in OSAS will require race- and gender-specific validation. Therefore, the MCFSI may require modification to avoid any intrinsic bias related to question formulation or cultural sensitivity. Also, differences in education level, which were not collected in the study cohort, could influence MCFSI scores. It is certainly possible that race, gender and educational level could have an actual impact on cognitive function. There was no statistically significant difference in the AHI/RDI between patients with normal versus abnormal MCFSI scores. The MCFSI scores were collected on patients with OSAS who were already on CPAP treatment, hence any correlation between MCFSI scores and AHI/RDI could not be established in this study. However, it is possible that patients with a higher AHI/RDI were more likely to use CPAP (as seen by the higher AHI/RDI in the CPAP-compliant group) and thus more likely to have normal MCFSI scores, as seen by the trend for higher AHI/RDI in the group with normal MCFSI scores. Conversely, it is also possible that the higher level of CI in CPAP-noncompliant patients may itself have contributed to less use of CPAP, thus creating a vicious cycle of worsening CI and other sequelae of OSAS. There are several limitations to this study. These include a small cohort size and lack of formal neuropsychological testing. We did not have enough power to establish the strength of individual risk factors in the causation of CI in OSAS subjects. Data on the oxygen desaturation index were not collected. Also, the relationship between OSAS and cognitive function may not be linear (possibly because of complex relationships between risk factors and disease). Future large studies are needed to further evaluate the association between OSAS and CI, the pathophysiological mechanisms underlying the CI, potential race/gender associations and targeted strategies to prevent/reverse the process. Further studies to evaluate the relationships between severity, gender and race are necessary. To the best of our knowledge, our study is the first one to use the MCFSI questionnaire for evaluation of cognitive function in patients with OSAS. Larger studies are needed to further validate the utility of this quick and simple screening tool for assessment of cognitive function in the OSAS population.
218
CONCLUSIONS CPAP compliance may improve the CI associated with OSAS. Thus, prompt diagnosis and treatment of OSAS and encouragement of CPAP use in patients with OSAS are critical. CI in OSAS is influenced by several factors and these relationships need to be evaluated further in future large studies. REFERENCES 1. Young T, Palta M, Dempsey J, et al. The occurrence of sleepdisordered breathing among middle-aged adults. N Engl J Med 1993; 328:1230–5. 2. Fava C, Montagnana M, Favaloro EJ, et al. Obstructive sleep apnea syndrome and cardiovascular diseases. Semin Thromb Hemost 2011;37: 280–97. 3. Fusetti M, Fioretti AB, Valenti M, et al. Cardiovascular and metabolic comorbidities in patients with obstructive sleep apnoea syndrome. Acta Otorhinolaryngol Ital 2012;32:320–5. 4. Lal C, Strange C, Bachman D. Neurocognitive impairment in obstructive sleep apnea. Chest 2012;141:1601–10. 5. Findley L, Unverzagt M, Guchu R, et al. Vigilance and automobile accidents in patients with sleep apnea or narcolepsy. Chest 1995;108: 619–24. 6. Naismith S, Winter V, Gotsopoulos H, et al. Neurobehavioral functioning in obstructive sleep apnea: differential effects of sleep quality, hypoxemia and subjective sleepiness. J Clin Exp Neuropsychol 2004; 26:43–54. 7. Wallace A, Bucks RS. Memory and obstructive sleep apnea: a metaanalysis. Sleep 2013;36:203–20. 8. Kim SJ, Lee JH, Lee DY, et al. Neurocognitive dysfunction associated with sleep quality and sleep apnea in patients with mild cognitive impairment. Am J Geriatr Psychiatry 2011;19:374–81. 9. Walsh SP, Raman R, Jones KB, et al. ADCS Prevention Instrument Project: the Mail-In Cognitive Function Screening Instrument (MCFSI). Alzheimer Dis Assoc Disord 2006;20:S170–8. 10. Canessa N, Castronovo V, Cappa SF, et al. Obstructive sleep apnea: brain structural changes and neurocognitive function before and after treatment. Am J Respir Crit Care Med 2011;183:1419–26.
Volume 348, Number 3, September 2014
