Abstract
Background
In the United States, 50 to 70 million adults have a sleep disorder. 1 Insomnia and sleep apnea affect 6% to 10% and 10% to 25% of adults, respectively. 2 Population 3 and primary car 4 studies report associations between disordered sleep and medical conditions such as obesity, diabetes, and chronic pain, and overall mortality. Primary care providers can play a critical role in identifying and reducing disordered sleep. 2 However, primary care settings typically do not routinely screen for sleep disorders, which may contribute to underidentification.5-7
Ethnically diverse urban primary care settings may be particularly important targets for sleep disorder screening. Low socioeconomic status and urban living are risk factors for sleep disorders.8-10 People who identify as black or Hispanic are less likely to report sleeping the 7 to 9 hours per night recommended by the National Sleep Foundation 11 compared to white, non-Hispanic in population-based studies. 12 Sleep disorders are underidentified in ethnically diverse populations; for example, only 1.3% of individuals in the Hispanic Community Health Study reported a diagnosis of sleep-disordered breathing, despite almost 40% meeting diagnostic criteria. 13
Few studies have evaluated sleep problems in urban primary care settings; approximately half of participants in each study identified as white, and half as black.4,10 These studies demonstrated high rates of poor sleep quality (78%) and elevated risk for sleep apnea (59%). 4 Identifying as black (vs white) was significantly associated with poorer sleep quality. 10
The current pilot observational study aimed to provide a preliminary evaluation of the prevalence of sleep problems, associations between sleep problems and relevant medical comorbidities, and primary care physician assessment of sleep screening utility in a predominantly black and Hispanic urban primary care setting.
Methods
Participants
Participants (N = 95) were recruited from waiting rooms at 2 primary care clinics affiliated with a major metropolitan hospital in the Bronx, New York. Inclusion criteria were the following: age ≥18 years, English or Spanish speaking, seeing a primary care provider, and capacity to consent.
Procedures
Graduate psychology students approached all patients presenting for primary care during collection periods in July-August, 2015 to screen for inclusion criteria and ascertain interest in participating. After giving oral consent, participants completed surveys in English or Spanish in clinic waiting rooms. Participants received $5 compensation. Providers received screening results prior to appointments, and rated screening utility after. The Albert Einstein College of Medicine Institutional Review Board (# 2014-3350) approved the study.
Measures
Demographics
Participants reported age, gender (male, female, other), Ethnicity (Hispanic, non-Hispanic), and race (white, black/African American, Native American, Asian, Pacific Islander, other).
Help With Sleep Problems
Participants responded “yes” or “no” to: “Have you ever tried to get help for your sleep problems?” and “Are you getting help for your sleep problems now?”; if yes, participants indicated treatment type. Participants responded “yes” or “no” to “Would you like to get help with your sleep?”
Comorbidities
Participants self-reported ever receiving a physician diagnosis of cardiovascular disease, diabetes, chronic pain, or migraine. Height and weight were obtained from the medical record to assess obesity.
Sleep Behavior
Items from the Pittsburgh Sleep Quality Index, 14 a well-validated measure of sleep quality, evaluated bedtime, latency to sleep onset (in minutes), rise time, and estimated hours of sleep per night.
The Insomnia Severity Index
The Insomnia Severity Index 15 (ISI) is a 7-item self-report insomnia screener. Responses range from 0 (“no problem”) to 4 (“very severe problem”). Total scores are classified as absence of insomnia (0-7), subthreshold insomnia (8-14), clinical insomnia (moderate) (15-21), or clinical insomnia (severe) (22-28). The ISI is a well-validated measure with excellent internal consistency (α = .90) and good sensitivity (0.86) and specificity (0.87) for identifying insomnia. 15 The ISI has also been validated for use in primary care, and has demonstrated excellent internal consistency (α = .92) in this setting. 16
The Berlin Questionnaire
The Berlin Questionnaire 17 is a 10-item obstructive sleep apnea screener. The Berlin Questionnaire assesses 3 categories: snoring behavior, daytime sleepiness, and history of hypertension and obesity. Individuals are deemed “high risk” for sleep apnea if they meet a minimum threshold in 2 or more categories and “low risk” if one or zero categories are endorsed. The Berlin Questionnaire is well-validated in primary care, and has demonstrated excellent internal consistency (α = .86-.92) and good sensitivity (0.86) and specificity (0.77). 17
Physician Assessment
Physicians were provided with screening results for each participant prior to the appointment. After the appointment, physicians ranked the perceived utility of the screening results, how the screening results informed patient care, and the extent to which sleep impacted the patient’s other health conditions.
Data Analysis Plan
Data were analyzed using SPSS Version 22. Descriptive statistics were presented for demographic and sleep variables. Chi-squares evaluated relationships between race/ethnicity and sleep problems. Bivariate correlations tested associations between descriptive variables and sleep problems. Odds ratios assessed associations between weight classifications (obesity vs nonobesity) and sleep problems.
Results
Participants (76.8% women, 46.3% black, non-Hispanic and 38.9% Hispanic, mean ± SD age = 44.6 ± 16.8 years) reported sleeping 6.2 hours/night (SD = 1.7) with 31.8 minutes to sleep onset (SD = 37.3) (Table 1). On the ISI, 30 participants (31.6%) screened positive for clinical insomnia (severe=7.4%; moderate= 24.2%). On the Berlin Questionnaire, 40 participants (42.1%) screened “high risk” for sleep apnea.
Demographic and Sleep Descriptive Statistics.
Approximately one-third (32.6%) of participants reported sleeping the recommended 7 to 9 hours per night (Figure 1). 17 A large proportion of participants wanted help with sleep (39.4%); only 12.6% of participants reported currently receiving sleep health care, and 28.4% reported previously receiving sleep health care.
Participants’ reported hours of sleep per night with National Sleep Foundation guidelines.
Self-reported medical diagnoses included high blood pressure (46.3%), cardiovascular disease (12.6%), diabetes (24.2%), chronic pain (14.7%), and migraine (12.6%). Chronic pain diagnoses were more common among persons with versus without clinically significant insomnia (57% vs 27%, χ2 = 4.97, P = .03). No other self-reported diagnoses were associated with sleep.
Mean body mass index (BMI) was 31.80 kg/m2 (SD = 7.20); 53.9% of participants met criteria for obesity. Weight status was not associated with overall insomnia severity; however, individuals with obesity reported fewer hours of sleep per night (mean = 5.79 hours, SD = 1.59) compared with individuals without obesity (mean = 6.57 hours, SD = 1.77), t = 2.19 (87), P = .03. Compared to non-obesity, obesity was significantly associated with high risk for sleep apnea (OR = 3.11, 95% CI 1.28-7.50). BMI was significantly higher among individuals who reported problems with snoring (mean = 34.16 kg/m2, SD = 8.08) compared with those who did not (mean = 30.60 kg/m2, SD = 6.46), t = −2.26 (87), P = .03.
Fewer than half of providers returned surveys (38.9%). Among providers who completed surveys (n = 37), the majority reported screening tools were at least somewhat useful (74.4%), and brought sleep problems to their attention (51.4%). Fewer questionnaires indicated that screeners informed patient counseling (17.1%), testing (5.7%), and referrals (5.7%).
Conclusion
This pilot observational study suggests insomnia and sleep apnea may represent an unmet need among urban, predominantly black and Hispanic primary care patients. Only one-third of participants reported sleeping the recommended 7 to 9 hours each night. 11 Approximately one-third of participants screened positive for clinical insomnia, and more than 40% were at high risk for sleep apnea; however, only 12% were currently receiving medical care for their sleep problems. In this sample, clinical insomnia was associated with chronic pain, while risk for sleep apnea was associated with obesity. These results extend previous research, where 59% of predominantly white and black primary care patients were at high risk for sleep apnea, and poor sleep quality was related with obesity. 4
Results from provider surveys suggest logistic constraints may limit the utility of routine sleep screening in busy urban primary care settings. Fewer than half of providers returned surveys, highlighting the fast pace and high clinical load of urban primary cares. The majority of providers indicated the sleep screeners successfully brought sleep to their attention. However, primary care providers may need improved internal and referral resources to act on information gained from sleep screening. Home sleep testing represents a valid and accessible alternative to a facility-based sleep study to provide diagnostic information regarding sleep apnea. 18 Integration of health psychology in primary care can improve access to evidence-based behavioral insomnia treatment 19 ; mobile applications are an effective alternative when behavioral health access is limited. 19
To address gaps in the literature, this study targeted urban primary care settings with predominantly black and Hispanic patients. This study addressed linguistic diversity with questionnaires in English and Spanish. Given the paucity of research in this population, these results provide new information about an understudied and important patient population; however, results may not generalize to other populations. This study evaluated sleep and medical comorbidities primarily through the self-report of patients who actively consented to participate in the study; future studies should utilize physician diagnosis or medical record to evaluate clinical diagnoses. Fewer than half of provider questionnaires were returned; provider results should be interpreted with caution. Future qualitative studies should evaluate barriers and facilitators to sleep disorder treatment in this population, and develop and implement individual- and system-level interventions to address identified gaps in sleep management in this population.
Footnotes
Acknowledgements
We would like to acknowledge the work of Claudia Lechuga, Alexandra Singer, Amy Grinberg, Carlos Marquez, Melody Willoughby, and Zarine Patel as research staff on this study. We would also like to acknowledge the generosity and contribution of the staff and patients at the Williamsbridge Family Practice and Family Health Center.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: EKS receives research support from the National Institute of Health (1K23NS096107-01 [Principal Investigator: Seng]) and the International Headache Academy (Principal Investigator: Seng); she serves as a consultant or has received honoraria or travel funding from the American Academy of Neurology, American Psychological Association Commission on Accreditation, and GlaxoSmithKline. CC, JL, TO, SS, MDM, and KAB have no conflicting or competing interests.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this study was provided by the Ferkauf Graduate School of Psychology at Yeshiva University to EKS.